the design i've draw in the previous post, took the simplest design when using a ground loop in a phone like the sensation. As you can see,
in reality, things are a bit more complex. Since you have those pins already connected one to another and those pairs to the phone ground (usb metallic shield) you already have a ground loop inside the phone and you actually extend it when you place the back cover.
Without trying to be more technical about multilayer pcb design or ground planes on a pcb, let's see what can be done about the gsm antenna and it's performance.
1. a logical first step would be to identify the pins corresponding to this antenna. Try to make some sort of cable with one end that can be fitted over those pins on the phone and the other end to an external antenna (piece of metal, if needed, use a simple metallic spoon or fork). The cable has to have 2 smaller cables inside, one for each pin. Once you find where are the pins for the gsm antenna we can proceed further. You may want to try other means to identify this, it's just that you may need a cable later on this "guide" so i've suggested to have one ready.
2. it's kind of difficult to experiment with high frequency antenna designs since there are no hobby tools capable of measuring these kind of frequencies (in the 0.8 - 1.8 - 2.4 Ghz area) and neither their amptitude in a given circuit. Since the phone itself has some means to measure this, we'll .. aah, use the phone. Signal bars, dB meters, whatever you find relevant. Another thing i do when i experiment, is to .. oversize things for more visible and clearer results. Of course, if, let's say you have a broken old phone and want to see if it still works, you won't remove it's battery and plug it to a car battery - you will blow it up. That's not one of those things were you want to oversize. Helpfully, antennas are of different nature. So, when you have to decide if a particular type of antenna is helpful, make a bigger one just like it. If the bigger one works, and the smaller one doesn't - then the design works but it's a size problem. If the bigger one doesn't work, then surely the smaller one won't work either here so you shouldn't try to make one to fit in the phone.
3. So, you got your pins right and know what's the gsm ones. Let's talk about some of the most simplest designs. Since RF energy is invisible and.. without any tools, impossible to measure, i'll be making an analogy with magnets and magnetic forces because they are of a similar origin but you can observe the effect of magnetic forces applied to an object even if you don't see the actual magnetic fields.
So, here are the 2 examples we’ll be talking today.
As you can see in the picture, the first one is a simple metallic plate and the second one contains 2 plates, one smaller then the other (about ¼) separated and not touching each other. Below each design, you have a representation of what would you see if we were to look at them from their profile, with the antennas standing vertically. Let’s talk about how they work. In case of RF energy, you would have a RF field, a spatial area on witch these waves propagate. If you place a metallic object in this field, an amount of this field’s energy will be passed to the metallic object (the process is called “induction”). We can measure this since this process generates a bunch of free moving electrons inside our metallic object. The stronger the field or the larger the surface area -> we “aquire” more electrons. If the field is too strong, a lot of electrons are being induced, they kind of “rub” on each other and surrounding molecules (it’s called brownian effect) so we produce heat - a microwave oven works like this. If we still were to pump up a greater field and to induce more energy, we will produce more heat. If we still try to induce more energy, we will obtain nuclear fusion. If we find ourselves rather more insistent and still try, we will obtain energy-matter convesions (like einstein's e=mc2) and if we still try to induce more and more energy, the matter we create will collapse under it's own gravitational field, thus we'll obtain a small black hole and surely we'll get banned from XDA for this. But to scale things down a bit and still talk about some lonely electrons, based on the general definition of the electric current, our electrons constitute a small electric current induced in the metallic object by that RF field. In a phone, that electric current is what’s being filtered and amplified. All forms of RF fields work just like this, so in the phone’s case, gsm, wlan, bluetooth etc. Thus, we need to “aquire” (it’s not the most physically correct term.. I know, but I try to keep it simple) more electrons. That’s the role of any antenna. The first design works great in areas were you have a relatively good field strength (phone signal) since it offers a big contact surface with the electromagnetic waves. This however is not always enough in places where you have a weaker field since the longer the distance to the source, the more the waves dissipate over a greater surface and from the point of view of a single receiving device – loose energy. The second design is a hybrid between those satellite dishes you may see on top of some buildings and some high gain antenna designs use in wi-fi networking. Simply adding a second metallic plate to the larger one, changes things quite a bit. I’ll get back on it after another schematic, showing how these 2 designs receive those badly needed electrons.
So, you can see, the red square (we’ll call it emitter) and our antennas. In the first case, the red thing emits and the black thing (antenna, profile view) receives those red curved lines (field energy). Because of this, a bit of red “appears” inside the black thing.
The second case looks a bit more complicated and yep.. it is. Remember that some time ago, I’ve talked about comparing RF fields with magnetic fields for the sake of simplicity. Well.. it’s time to do that stuff.. Imagine the red thing as a strong magnet. You have 2 iron plates at some distance from the magnet. You also glue these to keep them secure and not attracted by the magnet. In this setup, we say that we’ve “placed the metallic things inside the magnetic field generated by the magnet” (or something like this). The metallic plates are subjected to the force of the magnetic field, we can observe this easily because we have to keep them secured in place, not to move near the magnet. It’s the equivalent of the RF field described above. BUT something extra is also happening. While being attracted to the magnet the plates also become temporarely magnetized. So, they, themselves, will attract other metallic objects. This means that they have created a magnetic field of their own. The smaller plate, closer to the magnet, produces a larger secondary field, it’s effects reaches the second larger plate and get’s added up to the magnet’s field. Thus, the larger plate, receives a greater magnetic field in this case as opposed to the first example were only one plate is being used. The larger plate will also produce a magnetic field of it’s own, it will also reach the smaller plate, it will be induced in it, then the smaller plate will “re-emit” it to the larger one … and it’s kind of complicated even here.
The fact is, and it’s important, if we get back to waves.. it can go the other way around. Some bunch of electrons moving inside a metal plate will also create a RF field. So, maybe it’s easier to understand now, it’s actually rather difficult to design antennas since they act in both ways. They receive a part of the RF field, but they also can emit at the same time, a part of the energy they just received. The second design incorporates this facts.
4. After all of this.. how do we link those phone pins to these plates? Well that’s simple. In the first case, we will solder of at least secure 2 wires to 2 opposing corners of that plate. Those will be linked to the phone’s pins (polarity doesn’t count here). In the second case, some experiments must be done. First connect the larger plate to the phone, just like in the first case. Observe if you have any improvements, try to also connect the smaller plate to one corner or the other (by another wire). See what produces best results.
5. How big is .. big? When experimenting with larger antennas, I suggest that the larger plate to be around 50cm – 80cm in diagonal.
6. Materials… My choice would be copper. Thickness is very less relevant. Since copper foil is harder to get, you may also use aluminum foil if you manage to secure wires on it. Copper can be found on electronics store as prefabricated “blank” PCB’s. They are pretty cheap also. You can easily solder a wire on them. If you get these things, go for a single plated one, not the ones that have both sides covered with copper.
7. If you find a design that works for you in the large scale, then will try to find a way to make it smaller and fit it the phone's back case. But if the larger one doesn't work, if any large antenna design doesn't improve things much, the smaller ones surely won't help either and the problem must be searched in other places (phone's firmware, hardware etc).
