Make external connector on a USB dongle

It's the tiny rectangular blue "chip" thing in the upper right hand corner of the photo labelled WTC. It's a "meandering wire" ceramic substrate monopole antenna. This isn't the exact make and model, but it's the same type:

formatting link

Unsolder the blue chip antenna. Use TWO soldering irons like a tweezer. The trace near the point labelled A1 is the center conductor of the coax. The large ground area nearby is to the shield. Keep the exposed center conductor of the coax *VERY* short.

You may need to add an isolating series capacitor if the antenna has DC applied to it from somewhere. I can't tell from here.

If you want to be crude, cut two pieces of stiff copper wire to 31.1mm long. Solder one to the hot lead (one end of L7). The other to ground on the opposite side of the board. Instant dipole. Not great but much better than the original chip antenna. Add a reflector and you have a possible dish antenna feed.

Reply to
Jeff Liebermann
Loading thread data ...

Better idea. Leave the chip antenna in place. Cut the trace near the top of FL1 as close to FL1 as possible. Attach the coax center conductor to the top lead of FL1. Connect the ground to the large ground area to the left of FL1. Keep the exposed center conductor

*VERY short. C53 apparently acts as a DC isolator from the PIN diode diversity switch chip (half way hanging out from under the shield).
Reply to
Jeff Liebermann

Of course, if you only want to have an antenna with better send/receive characteristics, and highly directional would be OK, you could place the intact dongle inside a tin can and thus making a tin-cantenna.

formatting link
formatting link
Bob

Reply to
Bob Alston

Y'er right. It's C35.

Incidentally, John Navas wrote a page on build your own antennas:

formatting link
The monopole will work, but not as good as a dipole. The reason is not very subtle. The way the waveguide can antenna works is that the dipole is roughly 1/4 wave from the bottom of the can. The bottom acts as a reflector. When the signal from the dipole hits the bottom of the can it reverses phase and bounces back toward the dipole. When it hits, it's exactly 360 degrees (in phase) at 2.4GHz. That causes any signal at the dipole to be reinforced by the reflected wave bouncing off the bottom of the can. Reinforcement mean gain (a good thing).

However, if the bottom is not exactly the correct distance from the dipole feed, then one gets partial reinforcement or at worst, cancellation. That's the problem with a monopole versus a dipole. Half the signal is missing and there's no place for anything reflected off the bottom, on the ground side, to provide signal or gain. The monopole side of the USB adapter will work, but the antenna gain is that of an antenna half the diameter.

Drivel: With the common side mounted monopole, the reflecting bottom is not exactly 0.250000 wavelengths from the dipole feed. It's usually a bit farther. With a dipole, it will be closer to 0.25 wavelengths. In my non existent spare time, I've been tinkering on a proper model for the coffee can antenna:

formatting link
upon work by Ivor Hewitt.
formatting link
can change the monopole feed to a dipole or suspended monopole and see what happens. The spreadsheet still needs LOTS of work and checking and is nowhere near done. The results looks something like:
formatting link
the conical monopole feed. Volunteers and criticism graciously accepted.

Reply to
Jeff Liebermann

Thanks. That's the part by Walsin. I couldn't find the manufactory.

0 to 2dBi gain.

It's part of a Pi section matching network or low pass filter (or both). The RF path starts at the PIN diode diversity switch (half exposed from under the shield). It then goes to FL1, which is a

2.4GHz bandpass filter. C35 and C33 provide DC isolation. Usually, there's a u-FL test connector next, but it appears to be missing. The Pi matching network consists of C34, L7, and C33. Then, it goes to the 50 feed at the antenna. The Pi matching or lowpass network is probably there because the residual pad to ground capacitances and uncompensated lead inductances result in not quite a 50 ohm system. The lowpass function is also handy to help meed FCC spurious and harmonic content specs.

That was also my original suggestion. However, I wanted to take the long leads and the filter out of the circuit and connect as close to the bandpass filter as possible.

Good luck.

Reply to
Jeff Liebermann

Walsin Technology Corporation.

Reply to
Jeff Liebermann

Hi. I've a Zydas based USB WiFi 802.11g dongle.

formatting link
Chipset Zydas ZD1211
formatting link
I would make an external connector but I don't know where I can put the signal.

Where is the antenna? Where I can make the solder?

Thanks!

Reply to
Tazzulella

I'd say that was c35, but I might be reading upside down ;-)

What about a 35mm monopole sticking up from the board, at the top of FL1, poking into a cantenna? I suggest 35mm presuming that the shield seen in the photo would be soldered to the side of the can, leave 31.1mm exposed inside the can.

formatting link

Reply to
dold

"Jeff Liebermann" ha scritto nel messaggio news: snipped-for-privacy@4ax.com...

Or this:

formatting link
formatting link
formatting link
;)

I don't know... why there is also L7 and C3? I would to unsolder the bluechip, cut the waste area under the bluechip (it's not used!), put on the chassis an SMA-RP connector and connect it at L7+ground by RG316 (or RG196) teflon cable.

It seems a good idea...

Tomorrow I'll try... Thanks for support.

Reply to
Tazzulella

"Tazzulella" ha scritto nel messaggio news:43890236$0$8500$ snipped-for-privacy@news.tiscali.it...

formatting link
formatting link
What's the meaning of WTC?

Reply to
Tazzulella
[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]

Not my page -- I simply linked to it in the FAQ and added a few links rather than roll my own.

Reply to
John Navas

"Jeff Liebermann" ha scritto nel messaggio news: snipped-for-privacy@4ax.com...

Uhm... today I've realized this:

formatting link
Ok, it works, but I don't know why... with the previuos modify on my PCMCIA card, without external antenna the signal was ZERO. Now, with this dongle, without antenna (the bluechip) I've got signal... With antenna mounted (a simple rubber D-Link antenna) the signal is not stronger than without...

I'm ready with 3 attemps.

with antenna connected:

1) at the end of diversity (before the filter) 2) at the end of filter (cutting C3) 3) on the SW1 place (cutting L7 and C33)

Next week I'll receive a R&S powermeter 3GHz capable, and I'll to try in these 3 positions. I've also a friend of mine with a Anritsu antenna/power/swr analyzer. I'll tell you...

Thanks again for support: virtual coffe (italian espresso) payed for you.

Reply to
Tazzulella

Not great. I see a few problems.

  1. You destroyed most of the ground plane. Not a great idea but probably acceptable under the circumstances.
  2. The wire length between the center conductor and where the "cable" hits ground is way too long. Ground it to something near where the center conductor hits ground.
  3. Where's the shield wire on the coax on the connector?

If you don't wanna use a pigtail, you might try an SMA bulkhead connector, similar to what you have. However, I would file out an exact match for the outline of the base of the connector out of the ground plane you chopped off, with a short coax or direct connection to L7.

Without a 50 ohm termination or antenna, the traces on the board start to look like radiators and act as an antenna.

No clue. I've successfully butchered PCMCIA cards with pigtails. When they don't work, it's almost always a short at the coax end.

Nope. You really need the filter. It's not just for cleaning up the mess from the transmitter. It's also to keep junk out of the receiver.

That's worth a try. C3 is only there for DC isolation of the test conenctor, which isnt present.

Good luck soldering the tiny connector that will fit at SW1.

You'll probably get about the same power level from #2 and #3. #1 will yield about 2dB more (filter loss), but without the spec sheet, I'm guessing at the attenuation. I have a home made power meter made from an SMA conector, some chip resistors, a microwave diode, and a filter capacitor. Works nice.

Ooooh... $15,000 and up. Can I borrow it? I'll give it back. I promise. Maybe.

Otherwise... home made RF test equipment:

formatting link

Thanks.

Reply to
Jeff Liebermann

"Jeff Liebermann" ha scritto nel messaggio news: snipped-for-privacy@4ax.com...

Uhm, where is the ground plane? The waste area of board was clean... It's not "coppered" (I don't know the exact word, sorry).

Yeah, I have thought the same thing... Next try it's to rebuild ground connector to board.

It's around the dielectric, soldered on the connector: you don't see it because the picture was taken at 90°. I have more pictures, but not inside my computer: I'll show you soon.

Ok, next dongle... ;)

Sure...

Ok, I'm resolding it ;)

One moment... I don't wanna to put a connector at SW1 place... I wanna to solder the teflon cable at SW1 pins with shortest central connector.

My R&S was JUST arrived!

I'm intersted to your homemade...

Uhm... a LOT of coffee? Nice tool, but I don't have the know how to use it...

I'll take a look... Thanks again...

Reply to
Tazzulella

Oops. You're correct. I thought there was ground under the area that you removed. Incidentally, the official term is copper "clad".

This might be of interest:

formatting link
card, but same connector area. Note the VERY short leads on the tiny coax. That's required.

Reply to
Jeff Liebermann

"Jeff Liebermann" ha scritto nel messaggio news: snipped-for-privacy@4ax.com...

Time ago I've done this:

formatting link
After an hour I've modified the connection shortening the central cable and soldering an "arc" of wire between the 2 points of ground, reinforcing mechanical. Now it's closed with strongest glue and I cannot show you... I'm be able to works fine, don't worry... ;)

The Rohde & Schwarz NRS that I've received isn't be able to read WLAN powers... or I don't know how use it... at UHF and VHF frequencies it's sharp than my bird 43. I'll try with the Anritsu, soon, and I'll make your home made powermeter for Wlan.

Which powermeter you advice me to build? Thanks again

Reply to
Tazzulella

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.