So I have what may be a somewhat interesting problem. I’ve been trying to find the best way to design an antenna for a board I’m building which uses the RMF12B. However, everything I find is geared towards getting some crazy range measured in football fields. I would like the range on mine to be as small as possible, but omnidirectional. These boards will be used to send data from one glove to another when they’re being worn. They should never need to work further apart than you can spread your arms. The highest power I would need is enough to have your arms spread and have to communicate on opposite sides of your body. The smaller I can make the RF bubble created by them, the better. I’m planning to have something on the order of 30 of these operating in an auditorium sized room at once. I was looking at a “Splatch” antenna :http://www.antennafactor.com/resources/ … 433-sp.pdf
But I’m thinking that would still take up too much board space and may give me way more range than I really want. For reference, these boards, of which there will be one in each glove, will be less than 2 inches square hopefully. What would be the best way to go about what I’m trying to do?
Note too that the radio chip per se is from SiLabs - 4421. Microchip also rebrands/sells. Both of these companies have application notes too.
for super-short range, first config. the radio’s TX power to lowest. Then make a PCB antenna with per one of the data sheets, or just stick any old 1/10th wavelength wire into the enclosure. At short range as you speak, it’ll work.
Alright. I’ve read through the data sheets as well as a few tutorials for designing PCB antennas. They still don’t really answer my question though. I guess what I’m after is the smallest way to make a minimum antenna which will work. Can I just put a really fat trace in a spiral under the wireless module and connect it to the antenna pin, can I use a regular sized one, does it really really need to be the .25 inches away from every other trace? The antenna designs seem to assume there are 2 pins for the antenna when I only have one. I don’t know how to deal with that. The RFM12B data sheet focuses mainly on the antenna design for the 800 and 900 something mhz versions and not the 433 I have. I really don’t want to increase the size of the PCB by a lot to put the antenna on, and I also would really rather not solder wire to it and have to worry about it getting ripped off since this is a wearable device.
I second the advice from stevech. Antenna design is extremely complex, which is why you don’t find many answers to your questions by casual browsing. You’ll just have to experiment to find what works for your application. By the way, the “Splatch” is amazing.
The Splatch is about 1 inch wide and is soldered to one end of your pcb, so it extends the board by about 1/2 inch. It is a steal for $2.00 from Rentron. The board is supposed to have a ground plane but I didn’t bother. I get a range of about 50 feet with my setup (much better than the original piece of wire), but I’m using a different module than the RFM12B, at full power.
Ok so I’m trying to build for the Splatch now. I have a ground plane on the bottom layer that comes out .35 inches, and a .07" trace .3 inches long on the top layer for the antenna. The trace running to the “ANT” pin on the Splatch is the same width and property as all the others on the board, and there are no other traces within .5 inches of the ground plane or big antenna trace. The trace from the ANT pin on the RMF12B to the Splatch is routed without any considerations though. Does this sound about right?
I don’t like it. The antenna takes up more board space than the wireless module itself. Is there no more compact way to do that? Can I just attach a semi-sizable .4 inch spiral trace to the antenna pin and call it a day, or does it absolutely require some sort of special consideration?
Since I can’t find a schematic for it, I’ve just imitated it the best I think I can. .04 inch trace, .5 inches out, .75 inches to the left. Nothing else there. I think I’ll try that and see how it works, might add a solder pad also just in case it doesn’t, I’ll solder on a wire and hot glue it down. Any thoughts on this?
Narrow is better than wide. Try to avoid current paths that fold back on themselves, cancelling the B field. I would consider trying a ‘coat hanger antenna’ made from a coiled up mag wire soldered back to ground from the RF pin. Mess 'round with a few ‘designs’ and stake down the best one with hot glue or somesuch.
Self-reply update, again. I’ve decided to loosen my grip on low range, as I’ve come up with several very useful applications for the increased range that proper design will bring. The splatch, however, is still too huge. I’ve been looking everywhere for a 433mhz chip antenna that isn’t in the 3+ dollar range, which seems ridiculous to me. Much time spent on that. Then it hit me. The RMF12B is available in other frequencies. I’ll be sourcing some 915Mhz modules, as a 915Mhz chip antenna is much much easier to get ahold of.
and now I have a new set of questions. I’m still kind of fuzzy on antenna design, and I’m not sure exactly how to implement that design in my circuit. How do I make a 50ohm trace in Eagle? And do I need to put the entire board on a ground plane excepting right around the feed line and under the chip antenna, or just part of it? The entire test board seems to be a ground plane. Does the ground plane go on the top or bottom layer?
I really appreciate help with this, as I’m trying to learn how this works so I can deal with it on my own in the future. Information on this topic is just a bit hard to find in laymans terms…
The HopeRF modules are small and my use in 915MHz with a short wire antenna is 500 ft or so, from indoors to outdoors through a wall or two. Low bandwidth and data rate is the key to range, along with TX power and antenna size.
trouble with the chip antennas, and helicals, and spirals, is that their gain is negative (in dBi). Even a short wire antenna is 2dBi or so. 916MHz chip antennas are -4dBi or more gain; helicals are worse, at -6dBi or so.
Yes, but I don’t want to have any off-board components. Hence why I’m trying to do either a trace antenna or a chip antenna. These will be sold to people who shouldn’t be trusted with a cell phone, much less an exposed circuit board. anything that can be broken will be broken, and I don’t want to deal with people ripping antennas off. Hot gluing one on looks unprofessional. Even the backup battery for the clock will be non-removable and secured very well.
Probably not. It’ll be inside a thicker-than-average glove, inside a pocket made of elastic maybe. I might design an enclosure for it, but I doubt it. If I did, it would be very form-fitting and plastic. The idea is very low bulk.
look at the vendors’ specs - and several vendors. Beware some vendors’ claims.
A quick read of that vendor’s data sheet shows no H and V plots of antenna gain. There’s a worthless spec of 0.5dBi MAX gain. It’s just a cold hard fact that the chip antennas, and other less than 1/4 wavelength antennas have poor gain, high VSWR (wasted Tx power), and poor coupling. Again, size matters. These teeny antennas are fine if your range goal is modest and the transmitter power and bandwidth product are supportive. The lower the frequency, the worse it is!
As you may know, antenna gain is normally given in dBi which is gain relative to an ideal isotropic antenna (sphere).