433 MHz Coil Loaded Antenna

Intro: 433 MHz Coil Loaded Antenna

In my 433 MHz projects I have been using a cheap (0.70 cnts) pair of Tx/Rx
modules. I have mostly used the transmitter and that is actually fairly OK with just a simple 1/4 lambda antenna, but is open for improvement

The receiver however is a bit crappy: without antenna the reach is maybe no further than a meter, but even with a 1/4 lambda antenna it is marginally more, even with free Line of Sight.

For any serious project that involved receiving data it seemed I needed the much better (and more expensive) RXB8 receiver. But as said, also the reach of the transmitter could use a bit of improvement.

However, when mining the internet for a coil antenna (trying to improve on the lengthy 17.2 cm stick antenna) I came across a design of Ben Schueler, apparently once published in elektor magazine. A reference to Ben's pdf (back up) would suffice to build it, but so is my picture and I can add my experience with it as well.

It is a so called coil loaded design consisting of 0.6mm wire wrapped around a 2.5mm core.
The picture gives a clear description: a length of 25 cm wire should be enough. At the base it is 17 mm long. Then goes into 16 turns over a 2.5 mm diameter core (Ben advises to use 1.5mm²black installation wire for this. I just used a screwdriver)

The results with this antenna are very good. The distance (with the cheap receiver as well as the transmitter) that can be covered easily goes to 25 m with line of sight, but also in-house the distance will be increased reaching other rooms with concrete walls in between, were earlier 3 meters with line of sight would be pushing the limits already.

I am not the only one with this experience. Many people confirm to me that it dramatically increased the range of the cheap Tx/Rx pair, read the comments!

Comments

[mp2022]

I used this antenna, a straight rod antenna, and spring antenna. With the same distance to test, final and best one is the straight rod antenna. By the way the RF modules I used are FS1000A and RXB6. Anyone has the same result as me?

[diy_bloke]

Well, I only compared it to a 1/4 lambda wire antenna and found the coil loaded one the best. Not sure if straight rod antenna and straight wire antenna would be the same

[mjward1]

By tightly wrapping the coil, you are basically trying to create a tuned circuit with a higher Q factor than you need. ari.torpstrom has calculated this to be a tuned aerial for around 315MHz, which I have no doubt is true. As soon as you wind an aerial with a tight coil you introduce an inductor with close cross capacitive windings, in other words a high Q factor filter.

[KenC7]

So, how to fix it?

[mjward1]

Just use a single piece of wire, 17.3 cm long and make a single loop if required for physical constraints I calculated the aerial with a single turn on a 1cm former and its resonant frequency was 430MHz

[KenC7]

So you think this antenna design is worse than a straight quarter-wave antenna? Is there any way to do a coiled design like this, but change the lengths, coil diameter, or number of turns, and improve this design? For example the picture is a wearable I did, and the length of this antenna design is very helpful. It fits pretty well, with only a small part of the antenna wire protruding from the case. (I added the coil at right this morning to make it a dipole.) The way I wear this, a 17cm wire would be stabbing me in the throat. But the current design picks up signals from just under 200' away with no problems. Doesn't seem to pickup signals from farther than that, however. I added the dipole bit to try to improve that. A better tuned coiled antenna would be terrific, but a straight wire, not so much.

[mjward1]

17.3cm is the correct length of wire for a 1/4 wavelength straight antennae. Once you start to wrap the wire around a former, then the resonance of the aerial will drop in frequency, it wont stop working however as an aerial just the distance will be reduced. For the most efficient use you will want the aerial to receive around its resonance point. I just did a calculation for best nearest case. 10mm former with 10 turns with a total coil length of 30mm with 1mm wire (wire diameter won't make that much difference at these sizes. 0.6mm is just as good) Total physical wire length would be 316mm. Resonant frequency should be around 426 MHz

[KenC7]

So it's basically all coil then? I think I can make that fit - thanks! I'll try that out.

[SaifullahK5]

KenC7 .... Did you tried that 10 turn coil antenna design suggested by mjward1 ? If YES then was it a better success over the antenna design(coil loaded) as given in this instructable ? please share your experience ...

[KenC7]

Sorry, I haven't had a chance to try it. It was the middle of hunting season, and it wasn't a good time to make changes. I'll try to dig that up and see, but I don't have a good way to measure signal strength with that setup. Probably better if someone who can get a signal strength reading tries it out. I have other devices that do give an rssi, but that'll take even longer to rig up.

[portbox]

Just for everyone's knowledge, I tried all 3 antenna designs today. The first was a straight 17.3cm piece of wire, the second was the coiled wire in this instructable, and the third was mjward1's design. There is a picture attached. Not using fancy equipment, but simply moving the transmitter around my house and seeing which antennas worked and which didn't at difference distances was my testing method. The 17.3cm straight piece of wire performed the best, then the coiled one listed on this page, and mjward1's design performed the worst of the three. I'm not sure why. I even tried to stretch the total length of mjward1's design to see if a taller coil stack would help, but it didn't. I am going to stick with the straight wire even though it's long and ugly.

Edit: This was tested while plugged into a breadboard. Once I soldered the straight wire onto the protoboard, performance degraded considerably. Like really bad. Maybe interference with the rest of the Raspberry Pi.

[diy_bloke]

That is interesting. My results with the straight wire were pretty bad, that's why i went with the design as described here which immediately gave me much better results

[mjward1]

I think you may have not taken into account that an aerial needs a ground plane to reflect against. This is a quarter wavelength aerial that needs a metal plate or copper clad grounded PCB to act as the other half, making this a 1/2 wavelength dipole.

You also need to take into account the type of wire you are using, as the design was for a resin coated wire, not a plastic sleeved wire, which will detune the aerial and make it lose performance.

If it is not working then you can always try stretching the coil and cutting it shorter, or just use the straight wire, in either case, make sure the ground plane is there.

[portbox]

Yes, you are right. I did not consider a ground plane, I am pretty new to this and don't know anything about electrical engineering. That is also required for this instructable right? I don't see it mentioned anywhere. Is everyone that is using this design have it connected to a ground plane?

Regardless, how would you suggest I do this for my project? I am using a Raspberry Pi Zero W with this protoboard on top, and this RF receiver. The parts together look like the attached first image. The entire thing is maybe 2 inches long, not including the antenna, it's a small device. The soldered wires are connecting the chips ground, 5V, and data pin to the Raspberry Pi. If the chip is grounded to the Pi, is the Pi serving as the ground plane or no? Would I need to add another metal layer to this configuration? I appreciate your help.

[mjward1]

I can see the matrix PCB, which is not really good enough given all the holes would need to be soldered together to create a ground plane.
I have seen people use a metal plate, a copper clad board or even a metal box with the aerial going through a hole in the plate/box.
Theoretically if the pi is sat on a board with a ground plane, this may work, but given the fact that RF is not good for interference with data lines, you may want to add a ground plane over the PI anyhow.
I would opt for a piece of metal on spacers, or an alloy box, being the cheapest method, the aerial perpendicular to the ground plane

[KenC7]

Here's what I can add. I've done two different versions of (more-or-less) the same device. The one on the left is the pendant (necklace) version that I was working on before. The one on the right is a display version of basically the same thing. I wear the one on the left, and it uses a vibration code to indicate when a motion sensor some distance away (clear line of sight) has gone off. The one on the right sits on the floor of my raised hunting blind (walls are fabric), and gives me a letter/number combination for the same thing.

In both cases, the yellow oval is the antenna, and the amber rectangle is the 433mhz module. Notice in both cases, the antenna is perpendicular to the plane of the module's ground plane. Also, in both cases, the antenna ends up pretty much perfectly vertical. Both antennas are made of the same 22ga solid copper hookup wire.

Both of the antennas reliably pick up signals up to 600ft. away. I suspect they'd handle longer distances than that, but don't know. Also, I can't get a signal strength reading out of these - they're very dumb modules. I tried to make the antenna mjward1 suggested, but couldn't get it into my package. And since both of these work fine, I've just stuck with what I have.

I have two types of sensors. I have long range Olymbros sensors that use a professionally made SMA attached vertical-up antenna. For my home-made sensors, I use the straight wire antenna, and the wire is actually straight down. Both get picked up fine by the two receivers.

[portbox]

Did you create a metal ground plane or is there a layer in the module? I think I'm missing one in my device and that's why the receiver distance is so short. I need a reflective ground plane for my antenna, I think. My little device is only like 2 inches long.

[KenC7]

The modules I'm using have a ground plane on the backside of the PCB. I'll have to dig a bare one up and get a good picture. The oscillator and the few other components are mounted on top, and pretty much all of the back side is a single plane of metal. Just one short bit of separated trace on that layer.

[diy_bloke]

yes it is. goodluck

[0000007]

@ diy_bloke - > " yes it is " == mjward1`s design better than " coil loaded antenna "design ?