PDA

View Full Version : OT 390mhz antana



GKman
05-05-2018, 09:08 AM
I use a Chamberlain/Sears 390 mhz garage door opener. The transmitter range is 25 feet maximum. I extended the antenna wire outdoors with solid insulated 18 gage copper to even get this. Would like to try coax cable. A 1/4 wavelength antenna is 7.55" long. Though I would remove this much shield from the outdoor end for the antenna. Should the shield be grounded at the opener?

Benta
05-05-2018, 02:40 PM
Should the shield be grounded at the opener?

Yes.

Let us know how it works out.

Paul Alciatore
05-05-2018, 04:32 PM
Radio antennae, both for transmitting or for receiving, are a carefully engineered item. First I want to state that the transmitting and receiving antennae work exactly the same and are designed exactly the same, except for the possible power levels involved. You can transmit TV signals with a rabbit ear antenna, but not at a 50 K Watt level; it would self destruct with that much power/current. But, to illustrate that point of reciprocity, the people who make those high power antennae will often test them on a range by RECEIVING signals with them instead of transmitting.

Anyway, back to your obviously low power situation. A well designed antenna system has three elements: the amplifier circuit that provides the signal or receives it, a transmission line, and the antenna itself. For maximum efficiency these three elements should be designed to work together to bring a maximum amount of the signal (the power actually) from one end to the other. In your case, this means that the antenna picks up a small signal from your remote and that small signal has to be conducted to the amplifier at the other end, in the garage opener unit. This is the job of the transmission line. Now most garage door openers have a very short transmission line, perhaps only an inch or less long and this will be just a short wire. This is done for simplicity and low cost and they can get away with it because the length is much less than a wavelength. But when you try to extend the antenna any real distance from the opener, then you will need a real transmission line. To save drawing time I am going to use some drawings in this Wikipedia article:

https://en.wikipedia.org/wiki/Characteristic_impedance#Transmission_line_model

If you look at the second drawing you will see these three elements. It also shows the signal source at the left (your antenna) represented as an ideal Voltage source with a resistance (impedance) in series with it. It shows the transmission line and gives a characteristic impedance for it. Finally it shows a load or destination which is represented by a resistance (impedance) alone; that is your opener's circuit.

For maximum signal at the load (door opener) there is a mathematical relationship that the elements of the circuit must satisfy. I won't go into it, but this is the condition for maximum power transmission on a transmission line and it applies for all signals from DC and up to the highest frequencies. It says that the source and load impedances (those resistances shown) must be equal. There is a further calculation for high frequency circuits that states that the transmission line must have a "characteristic impedance" that matches those two source and load impedances. This characteristic impedance is not something you can measure with an Ohm meter. To make a long story short, these impedances; source, load, and characteristic impedance of the line must all be the same. This is not going to be the case with your present situation so there is probably room for a lot of improvement. That's the why. Here comes what I would do.

Most RF systems are designed around a 50 or 75 Ohm impedance. That means that all of those resistances would be the same and probably one of those numbers. I don't know what impedance your opener's circuit is designed for, but I would start by disconnecting all power from it (unplug it) and using an Ohm meter between the antenna terminal and the ground (case). If it reads 50 or 75 Ohms, you are in luck and can design with that value. If it reads a high resistance (1000 Ohms or higher), then it is also easy; you just add that resistance between that antenna terminal and ground. Notice that I said BETWEEN the antenna terminal and ground (the case), NOT in series with the connection to that antenna terminal. If it is an intermediate value, like several hundred Ohms, then post that number and I can calculate the size for that added resistor.

Now you need a coaxial cable. You can buy either 50 or 75 Ohm coax: RG-59 is 75 Ohms and is generally what is used in cable TV systems. Any cable TV coaxial cable will be a 75 Ohm one. RG-58 is 50 Ohms and can be purchased at electronic stores or on line. It uses the same F type connectors as used by cable installers for the RG-59: technically they will be slightly different for the different cable type but the same tools will work to install them. Given the choice I would choose the 75 Ohm cable because it is easier to find locally.

This coaxial cable's center conductor will be connected to the antenna terminal on the door opener and the shield will be connected to the ground (case). Keep these connections short.

The antenna should be designed for your frequency. You said that is 390 Mhz. Antennae are designed by frequency and a lot of work and math could be put into it. But here is a pretty good way to create one. Strip the end of the coax back by 1/4 wavelength.

Wavelength = 300,000,000 meters / frequency

Your wavelength = 300,000,000 / 390,000,000 = 0.769 meters = 30.28"

One quarter of that is 7.57" so you strip your coax back by 7.57". Now separate the shield from the center conductor with it's insulator still on it. One way to do this is to make an opening in a mesh shield at the point where it exits from the outer insulation and dragging the center conductor out of it there.

Now arrange the braided shield and the center conductor in a Tee in relation to the rest of the coax. You can use a rod of any insulating material to hold it in this position. The center conductor goes to one side and the braided shield goes to the opposite side. Fasten it together as you wish, but use only insulating materials, not conductors. If your coax has a foil shield and a "drain" wire, completely remove the foil and use the "drain wire in place of the braided shield. That is your antenna. It is not perfect, but it will be a lot better than what you have now.

If you are wondering about that source (antenna) impedance, it is not something that you must add. The antenna itself, as described here, will have a characteristic impedance at the operation frequency (390 Mhz) and it will be in the range of 50 to 75 Ohms. It can not be measured with a DC meter like an Ohm meter: it would require some complex and expensive equipment to actually measure. Or it can be calculated with some complicated math.

As for the transmission line, in theory it can be any length. But in practice, it will have some losses which will depend on the frequency: the higher the frequency and the longer the line, the greater the losses. So it is wise to keep it as short as possible. Another thing would be to mount the antenna in a location where there is a clear line of site between it and the transmitter. Tree leaves, bushes, or almost any other obstacle will decrease the signal.

I didn't have time to watch all of this, but I believe it illustrates this antenna technique.

https://www.youtube.com/watch?v=N-rUeIMYXbg

GKman
05-05-2018, 07:38 PM
Thanks Paul,
Quite an extensive explanation and I won't let it go to waste. Thought I had coax and of course didn't. Got RG6-U quad shield in hopes there would be something besides aluminum foil in it. It does, two layers of foil and two layers of braid. Braid doesn't want to take solder like it used to back in ought sixty nine but no surprise there. Found a coax connector male to male with a ground screw so I can connect copper conductors a couple inches long from the transmitter circuit board to it then a connector on the coax screwed onto it and I'm good to go. Resistance from antenna to ground measured infinite. Found a 68 ohm resistors (and others) I can solder across the antenna to circuit board ground if that's what you recommend.

J Tiers
05-05-2018, 11:17 PM
Just to stir the pot a bit more, the actual length of wire to make a tuned antenna is NOT the calculated wavelength.... the effective wavelength of a wire is less than the free space wavelength. So cutting to the free space wavelength will actually be a bit wrong.

You might guess at about 98% of the calculated wavelength and be closer. You do want the antenna to be resonant where the signal is, as that will give the best reception, which is what you need.

You will not have much of a way to determine how well your antenna is working, so the better you start out according to correct theory, the better your result will be. And the 2% error can actually make a good bit of difference in operation.

Frankly, in your place, I think I would just attach a longer antenna wire, and make it an odd number of quarter waves long. That is based on an assumption about the receiver which may be wrong.* If it does not do well, then change to a wire several half waves long. Don't bother with the coax. you probably do not need it.

* The assumption is that the receiver has a low impedance input. But if it had just a short wire antenna, it may have a high impedance input. With a low impedance input, then based on an odd number of quarter waves, you have a voltage maximum at the free end, and a current maximum at the receiver. But if the receiver input is high impedance, you want a voltage maximum at the receiver, so you want an even number of half-waves for the antenna.

The longer wire is to intercept more total signal energy.

Paul Alciatore
05-06-2018, 03:09 AM
The 68 Ohm value would be a compromise between 50 and 75 Ohms and it may work OK. It is closer to 75 Ohms so the RG-59 coax would be better.

My quickie design is no where near ideal and there will be impedance problems at the antenna end also so don't get overly worried about it. Try it and see if it helps. If not, resistors are cheap, $0.02 in hundreds and perhaps $0.25 for a couple of them.

As for connecting it, I can't see your unit so, solder is ideal. But you probably need to do that on the inside. Screw terminals are just fine and it will be indoors so there should be no problems from the weather. In either case, It is good RF practice to keep the leads as short as possible.

On soldering the coax braid, most braid is copper with no outer coating. So it can tarnish for a few inches from the cut end of a piece of cable. Often if you cut it back about six inches you will get to copper with less tarnish. Use a good electronic flux; they help a lot. And use a soldering iron with a high wattage rating. I often used my 100 Watt soldering gun for work on coax shields.

See more thought in my response to J's post.




Thanks Paul,
Quite an extensive explanation and I won't let it go to waste. Thought I had coax and of course didn't. Got RG6-U quad shield in hopes there would be something besides aluminum foil in it. It does, two layers of foil and two layers of braid. Braid doesn't want to take solder like it used to back in ought sixty nine but no surprise there. Found a coax connector male to male with a ground screw so I can connect copper conductors a couple inches long from the transmitter circuit board to it then a connector on the coax screwed onto it and I'm good to go. Resistance from antenna to ground measured infinite. Found a 68 ohm resistors (and others) I can solder across the antenna to circuit board ground if that's what you recommend.

Paul Alciatore
05-06-2018, 03:31 AM
J, In his original post he stated that he is trying to EXTEND the antenna OUTDOORS. I don't know about his door, but on most of them, mine included, that involves a distance of 10 to 15 feet or more just to get directly outside of the door. At 390 Mhz that is several wavelengths and is certainly enough to create standing waves if a proper transmission line is not used. With standing waves there is no way to know if the Voltage at the receiver end is at a high point or a low one. He could get a very low or even zero signal. This is probably what is happening now with his "18 gauge copper wire". I think he does need coax and he does need to make at least a stab at setting it up properly with at least a termination resistor at the receiver end.

As for that 98% thing, yes there are variations due to the type of wire, type of insulation, and other factors. But I doubt that his system is using exactly 390 Mhz either. It is probably some number around that. So there are other errors in the calculation that I posted above. Your 2% shorter length may be better or worse than the ideal number, neither of us knows for sure without checking the exact frequency.

One more thing that he can experiment with is the orientation of the antenna, vertical or horizontal or somewhere between. My door openers have short lengths of wire just hanging on the terminal board so they have vertical antennae. But I have no idea if that is the best or not.

I say YES to coax and YES to a termination resistor. And if you are going to get nit-picky about that 2% then get the exact frequency first. Your last paragraph is precisely talking about standing waves and that is NOT the way to design a good antenna system. You want to avoid standing waves.

As for determining how well it is working, there is a simple way. Walk away from the door/antenna and test the operation. See how far away it will operate the door. The further that is, the better the antenna is working. Simple and cheap.




Just to stir the pot a bit more, the actual length of wire to make a tuned antenna is NOT the calculated wavelength.... the effective wavelength of a wire is less than the free space wavelength. So cutting to the free space wavelength will actually be a bit wrong.

You might guess at about 98% of the calculated wavelength and be closer. You do want the antenna to be resonant where the signal is, as that will give the best reception, which is what you need.

You will not have much of a way to determine how well your antenna is working, so the better you start out according to correct theory, the better your result will be. And the 2% error can actually make a good bit of difference in operation.

Frankly, in your place, I think I would just attach a longer antenna wire, and make it an odd number of quarter waves long. That is based on an assumption about the receiver which may be wrong.* If it does not do well, then change to a wire several half waves long. Don't bother with the coax. you probably do not need it.

* The assumption is that the receiver has a low impedance input. But if it had just a short wire antenna, it may have a high impedance input. With a low impedance input, then based on an odd number of quarter waves, you have a voltage maximum at the free end, and a current maximum at the receiver. But if the receiver input is high impedance, you want a voltage maximum at the receiver, so you want an even number of half-waves for the antenna.

The longer wire is to intercept more total signal energy.

The Artful Bodger
05-06-2018, 06:45 AM
I believe this is one of those situations where too much theory will not advance things much as there are just too many unknowns, we do not know the exact frequency, we do not know the characteristics of the transmission line, we do not even know what we dont know. The answer is to suck it and see.

I suggest running coax to where you want the antenna to be keeping the last metre or so of the cable straight. Get a couple of lengths of heavy wire that will be stiff enough to stand as the antenna. Do a few tests for maximum range and if not far enough cut a quarter inch off the antenna and try again. The first few times you cut a piece off you may find the range actually decreases but somewhere before you get down to 7 inches or so the range should peak.

ironmonger
05-06-2018, 09:56 AM
<snip>

Most RF systems are designed around a 50 or 75 Ohm impedance. That means that all of those resistances would be the same and probably one of those numbers. I don't know what impedance your opener's circuit is designed for, but I would start by disconnecting all power from it (unplug it) and using an Ohm meter between the antenna terminal and the ground (case). If it reads 50 or 75 Ohms, you are in luck and can design with that value. If it reads a high resistance (1000 Ohms or higher), then it is also easy; you just add that resistance between that antenna terminal and ground. Notice that I said BETWEEN the antenna terminal and ground (the case), NOT in series with the connection to that antenna terminal. If it is an intermediate value, like several hundred Ohms, then post that number and I can calculate the size for that added resistor.

<<snip>>

Measuring the receivers output impedance in this manner will yield absolutely nothing of value.

The antenna circuit is a tuned circuit whose DC resistance value may be anywhere from a few tenths of an ohm to nearly an open circuit and has nothing to do with the RF impedance.

The designers of this receiver likely paid only lip service to input impedance as the circuit was designed to operate over a very shot distance and the attenuation of the transmitted signal was relatively unimportant.

The method that you describe for matching load values is not correct, and while may make the receiver happy the signal would be severely attenuated.

In the OP's case its easier to use a little empirical effort rather than a calculated one.

As a Amateur radio practitioner of some 50 odd years a little 'just try it' is often sufficient. I have successfully made radio contacts using carefully calculated and impedance matched antennas or a chunk of wire literally tossed into the trees with a computer controlled impedance matching system... the bottom line is can I communicate...

In response to the OP, the shield should be grounded at the receiver. I would try stripping the insulation back about 7" at the antenna end, pulling the center conductor out through the braid and pulling the two wires perpendicular to the coax in a "T" shape and see what happens. Or just try the bare wire...

In the words of a ham I know from New Zealand "let us know how you travel" which I assumed was a colloquialism for "let us know how it goes"... neither one of which makes a whole lot of literal sense. I guess that’s the definition of colloquialism :>)

J Tiers
05-06-2018, 09:59 AM
The usual receiver for these things is rudimentary, and the usual antenna is not in any way matched. It simply has nothing to do with proper RF practice in any way, it is intended to be cheap and more-or-less functional.

I specifically would not bother extending outdoors, unless the house is clad in sheet metal. Using a longer plain wire is simple, and generally works. Point being that the OP's idea may, IMO, be way overthinking and over complicating the problem.

As for the complaint of standing wave "design", Paul, you of all people ought to know that the idea described is merely making a tuned antenna and matching to the receiver input.

And you should know that going to extremes to create a 50 or 75 ohm system, only to throw away the received signal power in a matching resistor, is no better.

The OP does not have any test capability beyond basically pacing off the distance the thing works at, and comparing. The more variables you throw in, the harder it is to debug.

Putting a longer antenna wire on, with some idea of the correct length makes sense and is one variable. Snipping off a bit at a time to find the length that works best is equally simple.

Looks like Ironmonger and I were posting at the same time, but his idea may work too, if you really want to use coax.

Another one would be pulling the braid back down over the coax.... "bazooka" antenna http://www.hamuniverse.com/vertbazooka.html

Seastar
05-06-2018, 11:31 AM
Measuring the receivers output impedance in this manner will yield absolutely nothing of value.

The antenna circuit is a tuned circuit whose DC resistance value may be anywhere from a few tenths of an ohm to nearly an open circuit and has nothing to do with the RF impedance.

The designers of this receiver likely paid only lip service to input impedance as the circuit was designed to operate over a very shot distance and the attenuation of the transmitted signal was relatively unimportant.

The method that you describe for matching load values is not correct, and while may make the receiver happy the signal would be severely attenuated.

In the OP's case its easier to use a little empirical effort rather than a calculated one.

As a Amateur radio practitioner of some 50 odd years a little 'just try it' is often sufficient. I have successfully made radio contacts using carefully calculated and impedance matched antennas or a chunk of wire literally tossed into the trees with a computer controlled impedance matching system... the bottom line is can I communicate...

In response to the OP, the shield should be grounded at the receiver. I would try stripping the insulation back about 7" at the antenna end, pulling the center conductor out through the braid and pulling the two wires perpendicular to the coax in a "T" shape and see what happens. Or just try the bare wire...

In the words of a ham I know from New Zealand "let us know how you travel" which I assumed was a colloquialism for "let us know how it goes"... neither one of which makes a whole lot of literal sense. I guess that’s the definition of colloquialism :>)

Paul is correct!
He knows wher of he speaks.
Additionally, as a graduate engineer who has spent his life designing radio systems I think you are doomed to failure adding coax to the system to the "outside" of the garage. Unless it's a metal building it will probably decrease your range rather than increase it.
You would be better off adding a short length of wire to the existing antenna and then trimming it a 1/2 " at a time as described above till you get the max range.
If you insist on using coax ( bad idea) then make it an antenna by striping the cover off of the braid and pulling the braid back over the coax. That makes a sort of dipole antenna. The dimensions should be as described above and trimmed to max range.
Do not put a resistor at the receiver end of the coax or anyplace else. It will reduce your range by terminating the coax. My guess is that the input impedance to the receiver is somewhat close to the coax impedance. That's how I would design it and is common practice.
As Paul stated, a measurement with an ohm meter means nothing.
Try adding the short piece of wire and trimming it first. It's easy and may give you the results you need.
By the way a 25 foot range is not untipical.
Bill

GKman
05-06-2018, 12:29 PM
WOW. Original Poster Here. I followed Paul A's instructions to the letter and just paced off an ADDITIONAL 50 paces. (150' +) I'd have to climb a fence to test farther.
It is a wood frame building with steel garage doors. I have others in steel covered buildings with steel doors that work without any modification for some reason although the range is less than 50'. One is really short ranged and will add a cut and fit extension wire or coax extension to it also and report what happens.

Thanks again Paul A and all of the rest of you who contributed. I think there is magic, smoke, mirrors and JuJu involved so another idea may serve better in a different application. An old Ham radio enthusiast I worked for as a kid told how he fought troubleshooting an oscillator for months before he found out it only oscillated when his tester (coil capacitor and light setup) was near it. I'm happy with my monkey see-monkey do engineering in these black arts.

Seastar
05-06-2018, 04:10 PM
Congratulations!
I'm glad you got the range you need.
The steel door was probably part of the problem.
Bill

The Artful Bodger
05-06-2018, 04:17 PM
I can double the range of our garage door opener just by lifting the remote up to the level of the car windows.

Paul Alciatore
05-06-2018, 06:03 PM
Glad to hear it. And it seems to work gangbusters. Sounds like those steel doors may have been the problem. Can't explain the differences between that and other systems you have. But if it is working I would move on to other things.

Old truism in electronics: If you want an amplifier, design an oscillator: it will never oscillate no matter what you do. If you want an oscillator, design an amplifier: it will never stop oscillating, ever.




WOW. Original Poster Here. I followed Paul A's instructions to the letter and just paced off an ADDITIONAL 50 paces. (150' +) I'd have to climb a fence to test farther.
It is a wood frame building with steel garage doors. I have others in steel covered buildings with steel doors that work without any modification for some reason although the range is less than 50'. One is really short ranged and will add a cut and fit extension wire or coax extension to it also and report what happens.

Thanks again Paul A and all of the rest of you who contributed. I think there is magic, smoke, mirrors and JuJu involved so another idea may serve better in a different application. An old Ham radio enthusiast I worked for as a kid told how he fought troubleshooting an oscillator for months before he found out it only oscillated when his tester (coil capacitor and light setup) was near it. I'm happy with my monkey see-monkey do engineering in these black arts.

Paul Alciatore
05-06-2018, 06:35 PM
Yes, the input circuit of a receiver can have almost any impedance and it is a complex impedance. BUT, most RF front ends that I have seen have a fairly straight path from an antenna to the base of a transistor or the gate of an FET. The standard circuit for this is a RESISTOR network, probably with a fairly large value, series capacitor to block the DC path. This does not mean that the input impedance is strictly resistive, but that is not what I was concerned about. If that impedance was actually a high resistive value, then you do not have any real termination for the transmission line (coax). Using an Ohm meter which will make a DC measurement is not a positive way of determining that input impedance, but it will give you some information. If you do see a low resistance, then you do not want to add any additional termination resistors as the termination is already there. If it is an intermediate value, then you may need to add something. If it is a high resistance, like 5K or 10K Ohms, then the odds are that a termination resistor will help. BTW, this high impedance input type circuit is the simplest and probably the most used one in such devices and that is why I say "IF" and "the odds are". Why a proper termination resistor helps instead of just swamping down the signal is tied up in that maximum power transfer theorem that I discussed above and in the elimination of standing waves which J seems to think can help. RF is not DC and this is not a battery powering a light bulb. Standing waves makes it a crap shoot and I recommend against it. So a DC measurement of the Resistance is of some value if you know how to interpret it.

And you are technically correct when you say that, "The method that you describe for matching load values is not correct". The correct method would involve many thousands of dollars of complex test equipment and a large amount of skill in using it and interpreting the results. I have been there and done that. But I don't think it would be appropriate to suggest it to a poor guy who is just trying to improve a garage door opener. I gave him a way that would PROBABLY work and that was completely within the resources that he probably has. I am now retired and although I spent a 45+ year career in TV engineering, I no longer have any of that equipment. Generally, even in that area of TV engineering very few TV stations have such equipment and they hire outside consultants that do. Just that can cost over a thousand dollars a day. PS: in those 45+ years of working with all kinds of electronic equipment and with far more expensive test equipment at my fingertips, about 80% of my troubleshooting work was with a common VOM (Volt Ohm Meter). And although better meters were always present, most of that 80% could have been done with a $10 one.

And yes, the coax's braid should be grounded at the receiver. I clearly stated, "... and the shield will be connected to the ground (case)".

In any case, I am sure that other approaches could have been made to work. But my post was long enough.
And there are many, many ways of constructing an antenna at the end of that coax. Again, my post was long enough. And it apparently worked.




Measuring the receivers output impedance in this manner will yield absolutely nothing of value.

The antenna circuit is a tuned circuit whose DC resistance value may be anywhere from a few tenths of an ohm to nearly an open circuit and has nothing to do with the RF impedance.

The designers of this receiver likely paid only lip service to input impedance as the circuit was designed to operate over a very shot distance and the attenuation of the transmitted signal was relatively unimportant.

The method that you describe for matching load values is not correct, and while may make the receiver happy the signal would be severely attenuated.

In the OP's case its easier to use a little empirical effort rather than a calculated one.

As a Amateur radio practitioner of some 50 odd years a little 'just try it' is often sufficient. I have successfully made radio contacts using carefully calculated and impedance matched antennas or a chunk of wire literally tossed into the trees with a computer controlled impedance matching system... the bottom line is can I communicate...

In response to the OP, the shield should be grounded at the receiver. I would try stripping the insulation back about 7" at the antenna end, pulling the center conductor out through the braid and pulling the two wires perpendicular to the coax in a "T" shape and see what happens. Or just try the bare wire...

In the words of a ham I know from New Zealand "let us know how you travel" which I assumed was a colloquialism for "let us know how it goes"... neither one of which makes a whole lot of literal sense. I guess that’s the definition of colloquialism :>)

Paul Alciatore
05-06-2018, 06:57 PM
Sorry, but I couldn't let this go unchallenged. Yes, such receivers are very simple or as you say, "rudimentary". This is dictated by economics: when they can shave 1/10 of a cent on their cost, they do it. It only has to work, not be a proper design.

In my experience, just adding length to what is essentially a 1/4 wave antenna, is not very productive. It can help somewhat or it can just add to the problem. And he was going through a steel door which may not even have a window. Sounds like a great RF shield.

"And you should know that going to extremes to create a 50 or 75 ohm system, only to throw away the received signal power in a matching resistor, is no better." With a proper transmission line system, BOTH ends should present themselves to the coax as a matching resistance (no capacitive or inductive parts). This is what I was getting at in my long winded explanation. You can look up the math, I am not going to take the time to present it here, but the termination resistance of such a transmission line does not "throw away" any power. It MAXIMIZES the power that is transmitted to the receiving end of that line. So it is not wasting or throwing away any power: it is putting it exactly where we want it. If it is followed by an amplifier with a high input impedance, then things become Voltage dominated. Now you could get higher Voltages with standing waves, but standing waves also mean that multiple, successive cycles are being added together at that point and, with modulation, that can blur the information being transmitted (the door opening code). This could add to further difficulties in the proper operation and possibly reduce the range.

The original, OEM design of these openers does NOT have any real distance between the amplifier input and the antenna. So there is NO transmission line. But when you go to a distance of 15 or 25 feet, then you do have a transmission line, weather you like it or not, and it is better to do at least a rudimentary design for it.

Your "bazooka" antenna would also probably work, but in my experience a Tee arrangement, like I described, IMHO is a bit better.




The usual receiver for these things is rudimentary, and the usual antenna is not in any way matched. It simply has nothing to do with proper RF practice in any way, it is intended to be cheap and more-or-less functional.

I specifically would not bother extending outdoors, unless the house is clad in sheet metal. Using a longer plain wire is simple, and generally works. Point being that the OP's idea may, IMO, be way overthinking and over complicating the problem.

As for the complaint of standing wave "design", Paul, you of all people ought to know that the idea described is merely making a tuned antenna and matching to the receiver input.

And you should know that going to extremes to create a 50 or 75 ohm system, only to throw away the received signal power in a matching resistor, is no better.

The OP does not have any test capability beyond basically pacing off the distance the thing works at, and comparing. The more variables you throw in, the harder it is to debug.

Putting a longer antenna wire on, with some idea of the correct length makes sense and is one variable. Snipping off a bit at a time to find the length that works best is equally simple.

Looks like Ironmonger and I were posting at the same time, but his idea may work too, if you really want to use coax.

Another one would be pulling the braid back down over the coax.... "bazooka" antenna http://www.hamuniverse.com/vertbazooka.html

Paul Alciatore
05-06-2018, 07:08 PM
You may have spent a career designing them, but I spent a career making them WORK. I have had to make a lot of DESIGNS work in the field. I don't think I am the Paul you are saying is correct. But the OP says it did work. Not just work, but up to 150 feet. My door openers work to about 50 feet (to the street) and I would not have dreamed it would be that good. Three cheers for coax.

390 Mhz is not that high of a frequency and the losses in 20 or 30 feet of coax are there, but tolerable.

I don't have a circuit schematic or any specs on his unit, but I doubt that it has a low impedance input. I would guess 5K to 10K as that would be the cheapest way to go. And with just an antenna, no transmission line, in the original design considerations, it would be perfectly acceptable and would provide the strongest signal from the intended, makeshift antenna.




Paul is correct!
He knows wher of he speaks.
Additionally, as a graduate engineer who has spent his life designing radio systems I think you are doomed to failure adding coax to the system to the "outside" of the garage. Unless it's a metal building it will probably decrease your range rather than increase it.
You would be better off adding a short length of wire to the existing antenna and then trimming it a 1/2 " at a time as described above till you get the max range.
If you insist on using coax ( bad idea) then make it an antenna by striping the cover off of the braid and pulling the braid back over the coax. That makes a sort of dipole antenna. The dimensions should be as described above and trimmed to max range.
Do not put a resistor at the receiver end of the coax or anyplace else. It will reduce your range by terminating the coax. My guess is that the input impedance to the receiver is somewhat close to the coax impedance. That's how I would design it and is common practice.
As Paul stated, a measurement with an ohm meter means nothing.
Try adding the short piece of wire and trimming it first. It's easy and may give you the results you need.
By the way a 25 foot range is not untipical.
Bill

danlb
05-06-2018, 07:36 PM
Just a side note to the OP.

Many garage door openers are deliberately detuned to limit the usable range to right in front of the house. The idea is to add another minor obstacle to the thieves who drive through neighborhoods with transmitters designed to use rolling codes in order to open any nearby garage door. The further away from the house that the crooks can operate the more houses that they can hit at once.

Dan

The Artful Bodger
05-06-2018, 09:05 PM
Just a side note to the OP.

Many garage door openers are deliberately detuned to limit the usable range to right in front of the house. The idea is to add another minor obstacle to the thieves who drive through neighborhoods with transmitters designed to use rolling codes in order to open any nearby garage door. The further away from the house that the crooks can operate the more houses that they can hit at once.

Dan


Our garage door reliably triggers from a couple of hundred meters and hearing they are deliberately detuned is something I would expect to hear from an installer who cannot explain poor performance.

darryl
05-06-2018, 11:07 PM
There is what is called a velocity factor for various signal-conducting cables. For the typical RG6 I believe it's about .67, meaning that once you calculate a length for a particular frequency, you would cut it shorter to that extent. Looking at it another way, if the calculation shows an antenna segment length of 100 feet was right for the antenna element, then you could cut it to 67 feet instead. A few times I've made what's called a coaxial co-linear antenna, where you cut specific lengths of antenna elements, then reassemble them together (linearly) while reversing the shield and inner conductor at each rejoining. The end of the final section is shorted I believe. I made these for FM reception. Just hang them vertically and it acts like a multi-element antenna. For directionality you hang them using a steel pole, but spaced out in front of the pole by a certain amount. The pole becomes a reflector and you make it such that you can rotate the pole, just like the roof-top antennas of old.

The higher the frequency, the shorter the elements and probably the easier it would be to make and deploy. The result would be an antenna with a narrower but longer range. Even if you don't make a multi-section linear antenna but just use a single element (peeling back the braid for a specific length) you can still get a range advantage by placing it in front of a reflector by the calculated 1/4 or 1/2 wavelength. The reflector could be the aluminum siding, or a steel pole, or any other way of having a conductor parallel to and spaced away from the antenna. This could easily be a copper wire fastened to a 2x4, or you run the copper wire inside some small diameter pvc tubing which becomes the support for the antenna itself on a couple of stand-offs.

It seems you have solved the problem anyway, so that's what counts. No need to go to any great lengths if you don't need to.

J Tiers
05-06-2018, 11:47 PM
Paul A.... I will actually comment....

The antennas for those things are often just a length of wire that appears to have no particular tuned length. It may be picked specifically to limit reception, by NOT tuning it well

The added length of the wire can, and often does, help with getting more signal energy picked up. You just usually need to add tuned lengths. If the input is a high impedance, as may be the case, then you want a voltage maximum at the input, so you want a half wave or two half waves, etc... That high impedance would typically be intended for a non-tuned antenna,

If the antenna provided is a quarter wave, as you suggest, then you already know there is a voltage maximum at the open end, and a low Z point at the receiver input, which you want to maintain. So any longer wire needs to total up to be an odd number of quarter waves to maintain the same conditions.

The low Z input works fine if you are getting the intercepted signal power INTO the front end of the receiver. THEN it is actually used, and makes for better reception. Providing an external 50 ohm resistor will "match" a 50 ohm line to a high Z input, but the power that goes into the resistor is not going into the receiver. You may want the resistor in order to match the line and get predictable performance,.

Tuning the antenna may improve reception. And "tuning" may involve more than one quarter wave, although a quarter wave is the minimum typically used.

As for the metal door.... a door does not make an RF shielded enclosure. See comments about metal doors and metal buildings still receiving the signal. Shielding is something I have more than a little experience with, as we pre-tested all our products in RF tight rooms. The first was one I designed, and later we got a good Lindgren (sp?) room. Been there, seen it. If you WANT to block radio, you have to do a good bit more than put a big sheet of metal in the way (unless you are using microwave frequencies, that are shadowed effectively by fairly small door sized pieces).

The door MAY block direct transmission, but a door is not very effective for a fairly long wave such as the nearly one-meter length at 390 MHz, even if solid, which the doors rarely are. It will diffract around the door, and the opener is usually 2 or 3 m behind the door, and mounted near the top, or above, the closed position of the door.

A coax bazooka antenna, which I suggested, is reasonable and should allow extending the reception point away from the receiver box. But it is a good bit of complication. I'd first try the longer wire, and a bit of trimming, pacing off the reception distance as a comparison.

The most technically correct solution is not always the most effective for simple problems.

I do not think we have any fundamental disagreement here, it is just a matter of what gets the job done with the least fuss and extra expense.

danlb
05-07-2018, 02:50 PM
Our garage door reliably triggers from a couple of hundred meters and hearing they are deliberately detuned is something I would expect to hear from an installer who cannot explain poor performance.

Actually, it was discussed in a locksmith magazine years ago as part of the overall security plan for a residence.

danlb
05-07-2018, 02:59 PM
" The most technically correct solution is not always the most effective for simple problems." is a truism. It's also true that if you can approximate the technically correct solution you will spend a lot less time doing things wrong, breaking things, etc.

As for making antennas, the most basic designs are dead simple. 40 years ago I made a perfectly usable quarter wave dipole CB antenna from a piece of wire, coax and 20 feet of PVC pipe for a mast. My guide was a free RadioShack "how to" booklet meant for Hams. The math was simple. Tuning it was easy.

If I'd just used random lengths and tried to trim 1/8 inch at a time I'd still be climbing up and down the ladder.

J Tiers
05-07-2018, 03:37 PM
You want to get close first and be on the long side....of course.... it should not take long. Especially if you account for any added wire or trace lengths that are inside the unit/ on the PWB, etc, as you figure the length, so you are not chasing your tail as you start.

Many things are done by clipping to length, or other adjustments, bending, etc. When you do not have much if any test equipment, you pay for that in other ways. The OP has an opener and a transmitter, end of story. it may take a bit longer than checking with a counter and a grid-dip oscillator, etc (or better equipment).

Everyone is assuming that the antenna is tuned to begin with, That is not known, and may not be true.

The Artful Bodger
05-07-2018, 06:41 PM
Actually, it was discussed in a locksmith magazine years ago as part of the overall security plan for a residence.

If that is your concern best not to have a remote at all.

wombat2go
05-07-2018, 06:59 PM
The openers here have a lower range in winter. I suppose it is component temperature sensitivity in the receivers.
I was going to try better antennae than the receiver pig tails, but the wife fixed it first.
On a cold morning it is necessary to stand in the frame of the house back door and hold the transmitter under the chin.
It works perfectly.

bob_s
05-07-2018, 08:08 PM
The openers here have a lower range in winter. I suppose it is component temperature sensitivity in the receivers.
I was going to try better antennae than the receiver pig tails, but the wife fixed it first.
On a cold morning it is necessary to stand in the frame of the house back door and hold the transmitter under the chin.
It works perfectly.

Transmitter Button cells have VERY limited current output at low temperatures.

J Tiers
05-07-2018, 08:15 PM
Transmitter Button cells have VERY limited current output at low temperatures.

And things shrink, potentially changing lengths, resonance frequency, etc. It's a cheap consumer receiver.

GKman
05-09-2018, 01:04 PM
Pictures of project and one update. All of the shield of the coax at the antenna end is twisted together and enclosed in black heat shrink tube.
Tried to tidy up the installation with some insulated staples (shown) NO, tidy doesn't work. Range decreased to less than was before I started. Removed the staples from the ends and range is almost as good as the additional 150' I had gained.

https://drive.google.com/file/d/1JE7HuuE3fW25o1ht1jlAome-iYPPLcLK/view?usp=sharing

Paul Alciatore
05-09-2018, 01:15 PM
Blah, blah, blah! It works! Case closed so back to making chips.

Now if anyone knows how to get my "new and improved, all digital" cable service to stop cutting out, please tell me. Or better yet, tell the cable company as I am sure the problems are on their end, not mine.




Paul A.... I will actually comment....

The antennas for those things are often just a length of wire that appears to have no particular tuned length. It may be picked specifically to limit reception, by NOT tuning it well

The added length of the wire can, and often does, help with getting more signal energy picked up. You just usually need to add tuned lengths. If the input is a high impedance, as may be the case, then you want a voltage maximum at the input, so you want a half wave or two half waves, etc... That high impedance would typically be intended for a non-tuned antenna,

If the antenna provided is a quarter wave, as you suggest, then you already know there is a voltage maximum at the open end, and a low Z point at the receiver input, which you want to maintain. So any longer wire needs to total up to be an odd number of quarter waves to maintain the same conditions.

The low Z input works fine if you are getting the intercepted signal power INTO the front end of the receiver. THEN it is actually used, and makes for better reception. Providing an external 50 ohm resistor will "match" a 50 ohm line to a high Z input, but the power that goes into the resistor is not going into the receiver. You may want the resistor in order to match the line and get predictable performance,.

Tuning the antenna may improve reception. And "tuning" may involve more than one quarter wave, although a quarter wave is the minimum typically used.

As for the metal door.... a door does not make an RF shielded enclosure. See comments about metal doors and metal buildings still receiving the signal. Shielding is something I have more than a little experience with, as we pre-tested all our products in RF tight rooms. The first was one I designed, and later we got a good Lindgren (sp?) room. Been there, seen it. If you WANT to block radio, you have to do a good bit more than put a big sheet of metal in the way (unless you are using microwave frequencies, that are shadowed effectively by fairly small door sized pieces).

The door MAY block direct transmission, but a door is not very effective for a fairly long wave such as the nearly one-meter length at 390 MHz, even if solid, which the doors rarely are. It will diffract around the door, and the opener is usually 2 or 3 m behind the door, and mounted near the top, or above, the closed position of the door.

A coax bazooka antenna, which I suggested, is reasonable and should allow extending the reception point away from the receiver box. But it is a good bit of complication. I'd first try the longer wire, and a bit of trimming, pacing off the reception distance as a comparison.

The most technically correct solution is not always the most effective for simple problems.

I do not think we have any fundamental disagreement here, it is just a matter of what gets the job done with the least fuss and extra expense.

J Tiers
05-09-2018, 01:57 PM
Thanks, Paul... Always nice to see a post that does NOT quote someone for the purpose of making snarky comments about them. You are to be commended for NOT quoting and snarking.....

BTW, you do not know that this is the only solution that would work......