Joe McCauley wrote: >I'm looking for a cheap source for quarter wave whip antennae for use wi= th >this module >f> >This company do a suitable antenna at a rather pricey STG=A34.67. There = is >probably (I hope) a cheaper source for something which would do the job.= I'm >really looking for a flexible antenna of the type you see on walkie talk= ies. >Some rough handeling can be expected on these units. > >As I am somewhat uncertain when it comes to RF work, I am also looking f= or >some advice.... apparently a 1/4 wave whip antenna really needs a ground >plane. Does anyone know a rule of thumb for the relationship between the >antenna length (in this case ~16.5 centimeters) and the dimensions of th= e >ground plane? > =20 > A "perfect" ground plane is infinite, but generally for a 1/4 wave vertical, a 1/4 wave worth of ground plane in a circle around the antenna is "as close to ideal as is necessary". Ham Radio operators regularly stick magnetic-mount 1/4 wave VHF and UHF antennas to cookie sheets. Of course, virtually no portable devices have this much space at even UHF, and if the device is handheld or sitting on something, that stuff all affects the groundplane.=20 As far as materials go... I've seen plenty of devices that use stiff flexible wire as a vertical antenna. Do you really need the commerical rubber-duck? At RF, the transmitter sees no difference between a piece of wire and a fancy duck cut to the same length... so the question is all about packaging and whether or not someone is going to be dumb enough to pick up the unit by its antenna. (Most ducks have screw in connectors that can handle some of this type of abuse, and at least you can replace it if someone breaks it off... the wire would just rip from whatever strain relief you have provided and from the board.) Generally for a small electronic device, making sure the antenna itself is resonant is "enough". Obviously most RF devices are very unhappy if they're fed into something that's not a reasonable load and resonant so all the RF doesn't reflect back into the device. >I realise I've got a bit of reading to do, but am hoping not to have to >delve into the RF end too much. If I could get a range of 20 meters indo= ors >with this I'd be happy. The manufacturer quotes a range of 250 meters li= ne >of sight which one must assume is for ideal conditions. > =20 > 20m indoors (assuming no interference on your frequency from other devices in the structure and nothing wacky like metal walls) is easily "do-able" with anything that can get the RF out of the device and into the world.=20 >I am looking to network a large number of PICs in an indoor (factory typ= e) >environment using a multidrop serial protocol which already works in a >cabled situation. I want to change this to RF operation. > =20 > Factories have unique (and serious) problems at RF. Huge motors throwing RF noise, large stuff (and people) moving around into and out of antenna patterns constantly, large metal beams holding the building up that block RF very effectively at low power levels, supremely noisy power sources, etc. >All advice gratefully accepted. > =20 > Do some basic tests. Build a prototype and take a couple of them to various strange environments where you would expect RF blockage and get a baseline for how well they do. Nothing better than experience. Also, find some real environment where you would be hoping your devices would work, and put the manufacturer's RF engineers to the test... send them a description of the environment and see if they can provide you with real numbers and scenarios that would show the devices would work.=20 If your design is at all in control of any of the manufacturing process, you'll find out real quick if there are any safety issues, or if the RF device manufacturer will warranty them for use in that purpose! Line-of-sight numbers are window-dressing in the RF world... it's rare that you're really working with a perfect line-of-sight environment, ever= .=20 Anyone can take an RF output level, stuff it through the distance-squared rule and tell you if the receiver at the far end is sensitive enough to receive that signal, from an analog standpoint. And if the device is doing digital data, they should be able to provide you with a bit-error-rate chart for a particular signal-to-noise ratio. And of course, remember they're talking at RF line-of-sight... not visible light. Think about the materials your signals will have to pass through and put on a "mental filter" thinking about if you were able to see RF, what objects in the environment will soak it up. Where will it bounce?=20 Stuff like that. Sometimes it's easier to try it out in person, sometimes it's better to spend hours in the books coming up with theoretical numbers. It all depends on time, how you're getting paid, whether or not the system is only for one location or generic for any environment, etc. I recently had my butt kicked trying to get a relatively simple 802.11 2.4 GHz data link running in a building that had SPECIFICALLY been designed to keep RF out of certain areas... the building won. Getting a signal to a room in the hallway that was protected by screening in the walls, simply wasn't going to happen easily. We tried for a couple of days to come up with a solution that was also relatively out of the way. (Sure, I could have plopped the link radios on tripods and stuck them out in the middle of the large attached room pointed directly down the open-air hallway... but... that was a bit much.) If I would have had huge antennas and illegal 3w power amplifiers, I could have pulled it off... right through the wall... but it really wasn't worth it. Nate --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist