You can just use a piece of wire and save your coax in that case.
Exactly my point, and the reason this thread exists. This is the way it's done in the preassembled lights and how they said to do it for the replacement drivers they sent out for the first lights that had driver problems. Yet, I still don't know what the ground shield should attach to.
Should it go to ground? At the antenna end of the coax, both conductors are soldered to different spots on the same trace. Is that wrong too? Would grounding the shield at the driver end render the antenna useless? Well, more useless than if it used a plain non-shielded wire?
The stripped end you show in your first pic is the antenna the rest of the coax is just the feed line, the coax braid should go to the earth on the board, which I can’t tell in the pic. I’m not really into microwave stuff.
From what I know about antennas which isn’t much. The shielding is exactly what it says, shielding. It stops all interference from entering the inside wire where the signal is carried. If you where making a antenna with shielded wire you would strip the shielding because no signal can get to the inside wire, its being blocked by the shielding. I all so think it just goes to ground to absorb the unwanted signals which are picked up and carried by the sheilding.
My confusion comes from this: look at the pic in post#6. Up at the antenna end, the shield (drain) is soldered to the lower bar on the antenna, and the center conductor is soldered to the upper bar on the antenna. But the antenna is one contiguous piece of copper. I can't see how the center conductor & drain aren't carrying the same signal, and that grounding the drain at the driver end wouldn't just dump all the signal to ground.
If the drain were floating at the top end, then I'd have already put this together and would understand how it's supposed to work. I can see how the drain connected to ground at the bottom, and floating at the top, would dump unwanted interference to ground as a normal shielded cable does. As it is I cannot see how this setup will shield anything at all, and cannot see how the attachment method in the 'instructions' (both conductors glomped onto the same pad) can function any different than a plain unshielded single conductor wire.
I located JohnnyMac’s thread , and your right it doesn’t appear to be connected correctly. Though hell, it works & I’m certainly not 100%. RF has been called a black art, a lot of variables & difficulties to learn in order to really tune designs. Could look up the chip, maybe its above class 1, providing spare power to work even with the improper antenna connection.
RaceR86 posted a clearer picture of the pads (hope he doesn’t mind if I borrow it).
I marked the locations. All ext antennas I’ve seen are connected as such.
:beer: Thanks. OK, that's at least plausible.
The ‘shield’ doesn’t act as a traditional shield in many antennas.
Shielded wire is meant to carry a signal with out picking up any unwanted signal. When making a antenna no shielding is used or it couldn’t pick up any signal.
So the way they have it looks wrong, the green antenna picks up the signal plus any unwanted signals and is carried to the driver through the shielded wire.
The center wire should be all that is connected on the green antenna and at the driver signal input, the center wire would be the signal input with the shielded wire only connected to ground on the driver. To me that would be the correct way. The reason would be if the driver was giving off some kind of rf noise then the shielding would pick it up and send it to the signal input the way they have it. With the shielding isolated (from the center wire) running to just ground, it would be picking up the driver rf noise and running it to ground. No rf interference unless it comes from the remote green antenna.
.
Just my thoughts. 
Why the ‘shield’ is used like that I really don’t know. Open up a wifi device with a similar kapton pcb antenna on coax & your find it connected like that.
I’ve made some wifi diy yagi, biquad, waveguide antennas & the ‘shield’ is always used in a manner that seems counterintuitive.
You are confusing resistance and impedance. Resistance is for DC signals. Impedance is for AC/RF signals. Something that might have a DC resistance of 0 ohms and look like a dead short could have an AC impedance of say, 50 ohms. The higher the frequencies involved, the larger impedance effects come into play. In the microwave area, things get really tricky and critical. Something that would be unmeasureable at lower frequencies can totally block, short, screw up, etc a microwave signal.
So this is one of those "step 1: learn how to design microwave circuits, step 2: attach Asgard antenna" situations, huh? That damn 'how to draw an owl' picture again...
(Step 1: Solder antenna. (Step 2: have a drink. :bigsmile:
Will it be tuned 100% for the ext antenna, no, but it will work ok.
As said above, generally shielding isn’t used for wire antennas. At certain wavelengths of a given frequency a ground plane (not electrical ground) is used, for other wavelengths no ground plane is needed. I’d suggest looking at some DIY ham antenna websites . . some even have calculators for length if you know the frequency.
http://www.hamuniverse.com/antennas.html
EDIT: Note that shielding is used in cable of the proper impedance for transmitting the signal from transceiver to an external antenna (wire or otherwise).
for other wavelengths no ground plane is needed That is incorrect, there always needs to be some sort of ground plane needed except for ground independent antennas, half wave antennas for example.
Your missing something a wire antenna also depends on a good earth to work effectively, some keen antenna people will even use an earth mat buried in their yard.
No longer an issue, I tested it after each step during the process of fitting it into a one-off Maglite pill and it was fine, then after final assembly (drop in pill, tighten grub screw) it decided it no longer wanted to enter programming mode, so it had a little accident and is no longer of this world. I have better things to do. This crap is still in the pre-alpha stage and I have no idea how/why they are even still selling them.
Your example is exactly the fractional wavelength (for a given frequency) exception to which I was considering when I stated some relative wavelengths to a given radio frequency may not be ground plane dependent. For the OP, I’d be inclined to try for an even or half wavelength antenna (if you try again). Since they are probably small for the transceiver’s frequency, some precision and tuning is probably worthwhile (and quite possibly better for transceiver longevity).
A tower should always be well grounded to the earth for lightening protection, but towers are generally not designed to be a part of the antenna. Their earth ground still isn’t a part of the RF system. The example really doesn’t apply to a typical flashlight, unless perhaps it is being mounted in a fixed, outdoor location.
A counterpoise is essentially helping an antenna at a fractional wavelength for a given frequency to move toward a length with lower need for a ground plane (perhaps the most common being a 1/4 wave HT antenna needing but not having a good ground plane moving toward a 1/2 wave) by adding additional ground plane at the mount instead of adding length to a radiating element. Since the goal is likely to get the wire antenna out of the light, simply making the antenna a more useful length lessens or eliminates the need and complication of a counterpoise starting from another point.
Please take the EE-speak somewhere else if it doesn't also include relevant information about where to solder the two conductors of the existing, provided external antenna. I'm not going to redesign anything here, just put together the parts that came in the bag and let it ride. OK?
OK.
FYI those locations are not correct, the upper pad marked 'shield' is connected to something else on the driver, I think it's the output the bluetooth module uses to flash the LED to indicate programming mode. On version 2 I am going to leave the shield unconnected at the driver end, it may not be 'correct' and may reduce the range but it still at least works like that at an acceptable distance.