Post by thread:[EE] 434MHz aerial

I've been looking at various options for short-range transmissions at 434MHz. Less than 100m line-of-sight will do So far I've tried 1/2 wave (345mm), 1/4 (172mm) and 1/8 (86mm) straight aerials, using both copper and aluminium. 1/2 is best although 1/8 is still pretty good, apart from a couple of dead spots at 100m One application I'm evaluating would benefit from internal aerials. This could either be an 1/8 or a PCB trace. PCB trace would be preferable. My thought is to use a short link from the fixed-position user to a repeater. The repeater can have a 1/1 or 1/2 wave aerial to broadcast the longer range to the end receiver. This receiver should also have an unobtrusive aerial (PCB preferably), and I'm hoping that the power output of a repeater with its better gain aerial will be picked up more reliably with this internal aerial than from the user's short range transmitter I've Googled 434MHz aerial PCB and found this document, which has several 434MHz examples. The Bode plots are very helpful http://www.rfm.com/corp/appdata/antenna.pdf Opinions and suggestions please ? TIA Joe * * ********** Quality PIC programmers http://www.embedinc.com/products/index.ht

On 27 July 2010 21:34, ivp <spam_OUTjoecolquittTakeThisOuTclear.net.nz> wrote: {Quote hidden}> I've been looking at various options for short-range transmissions > at 434MHz. Less than 100m line-of-sight will do > > So far I've tried 1/2 wave (345mm), 1/4 (172mm) and 1/8 (86mm) > straight aerials, using both copper and aluminium. 1/2 is best although > 1/8 is still pretty good, apart from a couple of dead spots at 100m > > One application I'm evaluating would benefit from internal aerials. > This could either be an 1/8 or a PCB trace. PCB trace would be > preferable. My thought is to use a short link from the fixed-position > user to a repeater. The repeater can have a 1/1 or 1/2 wave aerial > to broadcast the longer range to the end receiver. This receiver > should also have an unobtrusive aerial (PCB preferably), and I'm > hoping that the power output of a repeater with its better gain aerial > will be picked up more reliably with this internal aerial than from the > user's short range transmitter > > I've Googled 434MHz aerial PCB and found this document, which > has several 434MHz examples. The Bode plots are very helpful > > http://www.rfm.com/corp/appdata/antenna.pdf > > Opinions and suggestions please ? > > TIA > > Joe > Joe, The pdf looks like it has plenty of good advice & examples. To get best range with low power you will need the antenna to be resonant at 434MHz and you will need to match it appropriately. A 1/4 wave whip on a ground plane will be close to 50 ohms at its base and its length can be trimmed for resonance. You'll need some equipment to get it perfect however - and then it will change with proximity to other things / people etc. (eg iphone 4) So a good starting point is to just cut to length as close as you can and see if it's good enough. Even a 1/4 wave long length of hookup wire may be enough, although probably not for 100m with the cheaper modules. Note that you need a metallic ground plane as the natural "earth" can be several wavelengths below ground level at this frequency. The feed impedance of the equipment may not be 50ohms either - although it's the best bet unless there is info to the contrary. Feeding a 1/2 or 1/8 wavelength antennae can be a bit more difficult as they will not be 50 ohms and possibly unbalanced. This can lead to additional losses & so reduce the actual transmitted and received power. This can be sorted out using baluns or transmission line stubs but would require suitable gear to fine tune & is likely to be more trouble than it's worth. A gain antenna (e.g a yagi) will give you longer range, but then again, may take a bit of setting up. If you know any hams you might be able to do some basic tests using simple gear (transmitter, receiver, VSWR meter) in the 70cm ham band. In NZ this is 430MHz - 440MHz (with some frequencies reserved for satellite use) so includes the 434MHz frequency of interest. I am in the position at the moment of working with RF gear and have access to test equipment. I measured one of my Wifi antennas and found it was slightly off frequency. A small adjustment to make it resonant and the performance has improved significantly so I can vouch for the fact that resonance does make a difference. R

> Feeding a 1/2 or 1/8 wavelength antennae can be a bit more difficult > as they will not be 50 ohms and possibly unbalanced. Well, a 1/4 wave whip is not balanced either. >This can lead to > additional losses & so reduce the actual transmitted and received > power. This can be sorted out using baluns or transmission line stubs > but would require suitable gear to fine tune & is likely to be more > trouble than it's worth. A better idea would probably be to use a 'rubber ducky' style coiled wire if a 1/4 wave is too long physically. In a lot of cases a 1/4 wave length of wire wound around a suitable diameter will do better than a straight shorter aerial. It may require a bit of tweaking to get the +/-j component tuned out, but with suitable test equipment that is easy to do. {Quote hidden}> A gain antenna (e.g a yagi) will give you longer range, but then > again, may take a bit of setting up. If you know any hams you might be > able to do some basic tests using simple gear (transmitter, receiver, > VSWR meter) in the 70cm ham band. In NZ this is 430MHz - 440MHz (with > some frequencies reserved for satellite use) so includes the 434MHz > frequency of interest. > > I am in the position at the moment of working with RF gear and have > access to test equipment. I measured one of my Wifi antennas and found > it was slightly off frequency. A small adjustment to make it resonant > and the performance has improved significantly so I can vouch for the > fact that resonance does make a difference. -- Scanned by iCritical.

alan.b.pearce@stfc.ac.uk wrote: > A better idea would probably be to use a 'rubber ducky' style coiled > wire if a 1/4 wave is too long physically. In a lot of cases a 1/4 > wave length of wire wound around a suitable diameter will do better > than a straight shorter aerial. It may require a bit of tweaking to > get the +/-j component tuned out, but with suitable test equipment > that is easy to do. In this case you might be able to use a plain old 1/2 wave center fed dipole. At 434MHz a 1/2 wavelength is about 13.6 inches, and you usually want just a little less than 1/2 wavelength anyway. A circuit board something like a 12 inch school ruler with the dipole along one edge and the electronics in a glob in the middle should work well enough. This should fit easily vertically on the performers' backs. If they are all vertical, you know the polarization, so a similar vertical dipole as a transmitter should work fine too. The more fancy antennas are there partly to be less dependent on polarization angle, but unless these performers are lying on the ground or bending over at the time the lights are switched, this actually works to your advantage. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000

> In this case you might be able to use a plain old 1/2 wave center > fed dipole. At 434MHz a 1/2 wavelength is about 13.6 inches, > and you usually want just a little less than 1/2 wavelength anyway. > A circuit board something like a 12 inch school ruler with the > dipole along one edge and the electronics in a glob in the middle > should work well enough. This should fit easily vertically on the > performers' backs Hi Olin. You've collided two threads. David Duffy was asking about communicating with performer's lights. My thread is about the design of compact aerials Since you mention a dipole, there's an example here (page 11) http://www.rfm.com/corp/appdata/antenna.pdf (106kB) of a PCB trace 'folded dipole' for 434MHz. In that form it's too long for the case I'd like to use, which is unfortunate because the author says it has very good performance. Perhaps it can be re- arranged to be wider and shorter to suit a small case. For example can the top half be rotated through 180 to be next to the lower half and still perform ? Where he notes 'feedpoint', does he mean that the signal is connected to just the lower trace end ? I'm guessing he does otherwise the two traces would be shown joined Joe * * ********** Quality PIC programmers http://www.embedinc.com/products/index.ht

ivp wrote: > Since you mention a dipole, there's an example here (page 11) > > http://www.rfm.com/corp/appdata/antenna.pdf (106kB) I just scrolled thru that document quickly mostly looking at the pictures, but it appears to be a nice writeup on compact antennas that can be integrated on a PC board. > of a PCB trace 'folded dipole' for 434MHz. In that form it's too > long for the case I'd like to use, which is unfortunate because the > author says it has very good performance. Perhaps it can be re- > arranged to be wider and shorter to suit a small case. Probably. That would probably also make the sensitivity in that plane more even. Even with fancy antenna modelling software, you still have to actually try it, then tune it for the performance you want. You can do some cool things with small antennas with maybe a L or C in there to resonate at your desired frequency or to null out the imaginary component of the impedence. Just don't expect a nice 50 or 300 ohms or something. At such small geometries, the impedences are usually higher. > For example > can the top half be rotated through 180 to be next to the lower > half and still perform ? Perhaps, but that would be a "major" change such as to be a different antenna. > Where he notes 'feedpoint', does he mean > that the signal is connected to just the lower trace end ? This is a symmetric self-contained antenna. It therefore is fed with a differential signal, just like a ordinary center fed dipole. Look at it carefully and you'll see it basically is a center fed dipole with the ends curled up. The antennas with ground plane only take a single ended feed because the voltage at the feed point is relative to the ground plane. Another way of saying this is that the antenna is not self contained since the ground plane is a integral part of the antenna. Actually all antennas have two feed points, but in some one of the feed points is implied because it is also the system's voltage reference (like a ground plane). There is really no such thing is a single feed point antenna if you look at the whole system the antenna is a part of. Hopefully this didn't confuse more than illuminate. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000

> Since you mention a dipole, there's an example here (page 11) > > http://www.rfm.com/corp/appdata/antenna.pdf (106kB) > > of a PCB trace 'folded dipole' for 434MHz. In that form it's too > long for the case I'd like to use, which is unfortunate because the > author says it has very good performance. Perhaps it can be re- > arranged to be wider and shorter to suit a small case. For example > can the top half be rotated through 180 to be next to the lower > half and still perform ? Where he notes 'feedpoint', does he mean > that the signal is connected to just the lower trace end ? I'm guessing > he does otherwise the two traces would be shown joined For an aerial like that it should really be fed from a balanced source. Think in terms of the old style VHF TV aerials with the ribbon cable coming down to the set. In that case the ribbon cable is carrying a balanced signal, about the earth connection, and goes into a balun inside the tuner to convert it to an unbalanced signal, i.e. one side is grounded. So for the traces pictured there the feed point really requires a balun that the transmitter unit feeds. A suitable balun is normally a few turns of wire on a ferrite core. The gauge of wire and core type both affect the characteristics of the balun. However it is worth getting down to your local library and having a look at an ARRL or RSGB handbook to see what the amateur radio guys do at 70cm for baluns, as there are tricks that can be done with coax or PCB traces to achieve similar results. -- Scanned by iCritical.

alan.b.pearce@stfc.ac.uk wrote: > A suitable balun is normally a few turns of wire on a ferrite core. > The gauge of wire and core type both affect the characteristics of the > balun. However it is worth getting down to your local library and > having a look at an ARRL or RSGB handbook to see what the amateur > radio guys do at 70cm for baluns, as there are tricks that can be > done with coax or PCB traces to achieve similar results. You can definately make your own baluns. Wrapping a differential pair around a ferrite core is a common trick. However, at his power levels and frequency you won't be able to make anything as small as what you can buy off the shelf. Various folks make baluns intended for exactly this purpose. They can have footprints down to about 1206 size. Coilcraft shurely makes some. I just checked, and in one 434MHz receiver I worked on we used a Mini-Circuits TC1-1-13M. There's no way you can make something that small by hand. The well defined specs in the datasheet would also be rather difficult to match.. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000

> Think in terms of the old style VHF TV aerials with the ribbon cable > coming down to the set That was the picture which sprung to mind when Olin said "differential" > A suitable balun is normally a few turns of wire on a ferrite core. > The gauge of wire and core type both affect the characteristics of > the balun I've some old 900MHz cellphone gear that I use for parts and can see quite a few small baluns on the boards, especially around entry and exit connectors. Although none of my applications are as big- business as cellphones I'll still make some effort to get the best out of the modules. I'll have a look at OTS small baluns The maximum permitted power level is 25mW BTW. The transmitter datasheet says its output is 3dBm typical at 3V, supply current 10mA However I do know that many of these modules will work up to 12V Vcc. For some reason the erroneous 3Vmax Vcc has been repeated and copied for years. I've had no failures with prolonged (years) 5V Joe * * ********** Quality PIC programmers http://www.embedinc.com/products/index.ht

Joe, Very roughly, Line of sight with no major reflections, X x power = sqrt(XX) x range. So 12.5 * power = sqrt(12.5) or about 3.5 x the range. (And yes, there will be exceptions to this). RP On 28 July 2010 10:48, ivp <.....joecolquittKILLspam@spam@clear.net.nz> wrote: {Quote hidden}>> The maximum permitted power level is 25mW BTW. The transmitter >> datasheet says its output is 3dBm typical at 3V, supply current 10mA > > I meant to add that 3dBm, according to > > http://www.referencedesigner.com/rfcal/cal_01.php > > is only 2mW, so there is quite some scope for more tx power, even > though these modules do pretty well at the target range anyway > > Joe > > * > * > ********** > Quality PIC programmers > http://www.embedinc.com/products/index.htm >

You might find the free program at http://www.eznec.com/demoinfo.htm useful for study and planning. I use the full version nearly every day in my antenna work (maybe better described as play). The demo version is entirely adequate for models in the uhf region and only a few elements. You can quickly plot patterns and compute gain (loss). You can build lab models and test them inside very easy using scraps of wire, cardboard, tape, etc. . A good field strength meter can be fashioned from a milivolt range meter, dipole (2- 1/4 wave wires) and a diode (germanium preferred, 1N34 typical). Copper is better than aluminum and easier to solder but aluminum is light and weathers better. Don't overlook using aluminum foil especially for experimental antennas! I won a contest a long way back by transmitting a video test pattern 12 miles on 30 miliwatts. The secret was the antenna - a 48 element Colinear Array at 40 feet... BTW, thanks for the reference to the paper, very interesting! John Ferrell W8CCW "For evil to flourish, all that is needed is for good people to do nothing." /Edmund Burke/ On 7/27/2010 5:34 AM, ivp wrote: {Quote hidden}> I've been looking at various options for short-range transmissions > at 434MHz. Less than 100m line-of-sight will do > > So far I've tried 1/2 wave (345mm), 1/4 (172mm) and 1/8 (86mm) > straight aerials, using both copper and aluminium. 1/2 is best although > 1/8 is still pretty good, apart from a couple of dead spots at 100m > > One application I'm evaluating would benefit from internal aerials. > This could either be an 1/8 or a PCB trace. PCB trace would be > preferable. My thought is to use a short link from the fixed-position > user to a repeater. The repeater can have a 1/1 or 1/2 wave aerial > to broadcast the longer range to the end receiver. This receiver > should also have an unobtrusive aerial (PCB preferably), and I'm > hoping that the power output of a repeater with its better gain aerial > will be picked up more reliably with this internal aerial than from the > user's short range transmitter > > I've Googled 434MHz aerial PCB and found this document, which > has several 434MHz examples. The Bode plots are very helpful > > http://www.rfm.com/corp/appdata/antenna.pdf > > Opinions and suggestions please ? > > TIA > > Joe > > * > * > ********** > Quality PIC programmers > http://www.embedinc.com/products/index.htm >

> the free program at http://www.eznec.com/demoinfo.htm > useful for study and planning. I use the full version nearly every day > in my antenna work (maybe better described as play) Thanks very much > A good field strength meter can be fashioned from a milivolt > range meter, dipole (2- 1/4 wave wires) and a diode > (germanium preferred, 1N34 typical) Back in the days when CB was popular I made an SWR meter to help a friend. Very much along the lines of what you describe, using a moving coil meter BTW. ISTR it work well > Copper is better than aluminum and easier to solder but aluminum > is light and weathers better The copper I have is heavy gauge laminated wire on a reel, which works but looks a bit scruffy when you try to straighten it out. The aluminium is TIG rod, similar gauge, nice and straight and shiny. I know it's probably prudent to use the best material but the appearance of aluminium does seem to carry some weight (no pun intended) Of course in critical or marginal applications efficiency would win out over aesthetics Joe * * ********** Quality PIC programmers http://www.embedinc.com/products/index.ht

> > It may require a bit of tweaking to get the > > +/-j component tuned out, but with suitable test equipment that is easy > > to do. > > > I would like to learn more. > > I have access to signal generators, spectrum analysers with tracking > generators and network analysers. > > I guess at resonance the load appears completely resistive and current > is in phase with voltage. > > Therefore, my first approach would be to use a CRO with a current > probe. Drive the antenna with signal generator at the desired > frequency and see how it behaves. > > However, I suspect you refer to another approach. > > I would welcome any suggestions. > > Justin Well, It is ages since I dealt with rf stuff, but my memories are that you would use the network analyser to generate a smith chart style display around the frequency you require, and then a suitable matching network would be used to tweak out the reactive components, and may be able to adjust the resistive value seen looking into the input side of the network to make it suitable for the tx or rx that you are using. My memories of working out Pi networks at HF type frequencies suggests that they can, with judicious component selection, be used to raise or lower the impedance to suit. -- Scanned by iCritical.

> Resonance center and pure resistive impedence are not always the same thing, > although usually close. The resonance peak is usually the easiest thing to > measure. For quick and dirty tuning it's probably enough. It is likely to be of very limited use in this application :-) - but I found long ago that best tuning on a dual '807 Class C final was achieved by maximising the blue glow which appeared on portions of the glass outside holes in the anode structure - caused by electrons hitting the glass, and not the blue inter electrode glow which is caused by gas. I had wondered if it was Cerenkov radiation, but Gargoyle reveals a number of people who say it fades with age of tube, suggesting some form of fluoresecence. Best results were very slightly off resonance. When your transistors start emitting Cerenkov radiation it's probably past time to leave the building. Russell

What kind of tube was this? Cherenkov radiation results from a particle such as an electron traveling faster than the speed of light in a medium. Since the index of refraction is the ratio of the vacuum speed of light to the speed of light in a specific medium, which for glass is in the 1.5 to 1.7 range, then the electrons would have to be traveling over 60% the speed of light with an energy of about 150 Kev.. It would be best to leave the building as the electrons striking heavy metal impurities in the glass would be showering you with X rays. I assume the applied voltage in the tube you used was not 150000 volts. :-) Pete RussellMc wrote: {Quote hidden}>> Resonance center and pure resistive impedence are not always the same thing, >> although usually close. The resonance peak is usually the easiest thing to >> measure. For quick and dirty tuning it's probably enough. >> > > It is likely to be of very limited use in this application :-) - but I > found long ago that best tuning on a dual '807 Class C final was > achieved by maximising the blue glow which appeared on portions of the > glass outside holes in the anode structure - caused by electrons > hitting the glass, and not the blue inter electrode glow which is > caused by gas. I had wondered if it was Cerenkov radiation, but > Gargoyle reveals a number of people who say it fades with age of tube, > suggesting some form of fluoresecence. > Best results were very slightly off resonance. > > When your transistors start emitting Cerenkov radiation it's probably > past time to leave the building. > > > > Russell > >

On Thu, Jul 29, 2010 at 8:05 AM, Olin Lathrop <.....olin_piclistKILLspam.....embedinc.com> wrote: > Justin Richards wrote: >> I guess at resonance the load appears completely resistive and current >> is in phase with voltage. > > Resonance center and pure resistive impedence are not always the same thing, > although usually close. The resonance peak is usually the easiest thing to > measure. For quick and dirty tuning it's probably enough. Olin, can you please elaborate? What criterion are you using to judge resonance here if not the absence of a reactive component? Sean > > > ******************************************************************** > Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products > (978) 742-9014. Gold level PIC consultants since 2000. >

Sean Breheny wrote: >> Resonance center and pure resistive impedence are not always the >> same thing, although usually close. The resonance peak is usually >> the easiest thing to measure. For quick and dirty tuning it's >> probably enough. > > Olin, can you please elaborate? What criterion are you using to judge > resonance here if not the absence of a reactive component? The easiest way to measure resonance is probably to feed it with a relatively high impedence source and then find the frequency that results in the highest amplitude. Unless you've got a very high Q antenna though, the resonance peak and the pure resistive point are going to be close enough that either is fine for tweaking things. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000

On Fri, Jul 30, 2010 at 5:52 PM, Olin Lathrop <EraseMEolin_piclistspam_OUTTakeThisOuTembedinc.com> wrote: > The easiest way to measure resonance is probably to feed it with a > relatively high impedence source and then find the frequency that results in > the highest amplitude. Unless you've got a very high Q antenna though, the > resonance peak and the pure resistive point are going to be close enough > that either is fine for tweaking things. I still don't understand, though. Aren't we talking about a dipole antenna? If so, it should look roughly like a series RLC circuit, in which case resonance is not the highest impedance point. When you say "amplitude" do you mean voltage across the antenna feedpoint terminals? Sean > > > ******************************************************************** > Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products > (978) 742-9014. Gold level PIC consultants since 2000. >