external Antena with moteino

[sid1202]

Hi, I have an application where the motino will by in a metal box. So I need an external ant. How do I connect that to the motino ?

Sid

[Felix]

RF coax to moteino board (GND + ANT port), out through a hole to your antenna. Depending on antenna you will need a certain length on your coax to keep the signals from bouncing in the cable.
See this thread for how to to it with dipoles.

[sid1202]

I've read the tread a few times. and for the life of me I can't figure out what the length of the coax changes.

I was my understanding that the coax isolates the signal between the output and the antenna.

Sid

[Felix]

It does not isolate at all, that would be very bad. It transports the signal, ideally without loss. But there's always loss. The length matters, that thread goes into more detail on that.

[sid1202]

That's what I meant. transport. Thee is loss but I assume that the loss becomes an issue with long runs of several feet. I would have at most 6" probably less than that.

Sid

[captcha]

Have you already determined what kind of antenna you need? And for what frequency?

[sid1202]

The frequency is 915 mHz,one side will have an external antenna since every this is metal and the other is a hand held plastic box with a sticky printed circuit antenna. Both antennas have a small lead with a uFl connector to connect to the board.

https://abra-electronics.com/wireless/wireless-antennas-en/1991-ada-slim-sticker-type-gsm-cellular-quad-band-antenna-3dbi-ufl.html

https://abra-electronics.com/wireless/wireless-antennas-en/ant-17-gsm-u.fl-connector-to-sma-antenna.html

[captcha]

The monopole antenna from your link only seems to have an SMA port. Just make sure the monopole is conntected to a metal ground plane. I'm assuming you're going to use an SMA-to-U.FL connector cable to the Moteino in a metal case. Then, both antennas will have an U.FL connector going to the Moteino. I guess the easiest would be to get some U.FL matching connectors and solder them to the Moteino as per Felix' post.

I've read the tread a few times. and for the life of me I can't figure out what the length of the coax changes.

Besides losses purely due to resistive losses (i.e. a long cable), the length of the coax becomes very important if the antenna does not have a 50 ohms feedpoint impedance (we call this a mismatch). Looking at the pages from your links it's impossible to tell whether these antennas are indeed that. What happens is that with a mismatch at the antenna feed point, some of the power going to the antenna is reflected back to the radio. Because the power through the feedline is AC (alternating current), the forward and reflected waveforms can either add or subtract. Imagine a point where the forward wave (from the radio) is at its highest, right at the point where a reflected wave is also at its highest. At that specific point the voltages add up. If that specific point is close the radio, it could damage it. This is why I recommend the use feedline of half-wave multiples so that the peak of the reflected waves are minimal.

What's working in your favour though, is that the coax that U.FL uses is pretty poor at handling high SWR and will result in resistive losses (warming up the coax). So in theory you could run a slightly higher SWR and get away with it.

If you're trying to get the furthest range and maximum efficiency out of your system you probably want to use different coax and antennas but I would just give it a try and see what it does. The radios on the Moteinos handle up to 3:1 SWR which is fairly robust but not unlimited. Never ever transmit without an antenna.

The PCB (sticker) antenna has a datasheet and on one of the graphs it shows that the SWR for 915MHz is roughly 2:1. For short runs I'd say that's acceptable.

[sid1202]

Thanks @captcha for the clarification. I understand better, not fully, but I get the point.

The monopole antenna is has a S.W.R <=2.0 and Output Impedance is 50Ω (stated in the specs section). it comes with a short cable that has a uFl connector and a SMA connector. The range is going to be short but fast. This is a wireless remote control setup for machinery. On the machine, the motino is going to be inside an aluminum box and the SMA port should have a good ground plane from the Al box.

Since both antenna's are 50Ω as the RFM69 radio and both have a S.W.R of around 2, I assume there should not be any resonance issue in the coax ?

[captcha]

both have a S.W.R of around 2

SWR is also a function of frequency and not just a property of the antenna.

An SWR of 2 (2 to 1, or usually written as 2:1) means that within a certain frequency range (they state from 824MHz to 960MHz) the SWR will be less than 2:1 (or so they claim). So just take the worst case scenario and assume it's 2:1. That should not give you any grief if those numbers are correct.

The range is going to be short but fast.

That's very open to interpretation. Are we talking cm's, metres? Same goes for 'fast', although you'll probably end up somewhere close to the speed of light.. 

I assume there should not be any resonance issue in the coax ?

Unless you create a trap (to create a multi-band antenna) there is no resonance in coax that we should worry about here. I think the term you're after is reflection. Unless the antenna is perfectly matched to the coax and the coax is perfectly matched to the radio there will be reflections. However, if you're trying to communicate with reasonably decent antennas (which these seem to be) over a range of say, 100m you should be fine.

[sid1202]

That's very open to interpretation. Are we talking cm's, metres? Same goes for 'fast', although you'll probably end up somewhere close to the speed of light.. 

lol your right. The distance will be generally within 5 meters and at a rates of 300 kbps.

And you're right I was thinking reflection.  So Ideally using a half wave length of coax would help. Since @ 915 mHz is only 6" I can tuck than in my box fairly easily.

I was reading your explanation and re read the thread on the subject and it reminded me of the wave combination effect on water.  one thing I'm curious about is if only the half wave length helps. What would a quarter length do? I'm not sure but I think it would also help, but not as much ?!?!?

My plan is to write state machines that will transmit state changes to one another. I plan to have a handshake every 500 ms when no state changes are present. This means that the radios will have a duty cycle that is very low most of the time with short bursts of busy times. With the info you gave me I'm pretty sure I'll be fine.

Thank you.

[captcha]

Ah, data rate, yes.. forgot about that..    Although I've never ran them that fast, 300kbps should be fine. Perhaps check one of the other posts on this forum about higher data rates and what to expect.

one thing I'm curious about is if only the half wave length helps. What would a quarter length do? I'm not sure but I think it would also help, but not as much ?!?!?

Just one quarter wavelength is probably the worst you can do. Again, if all is perfectly matched, then no issues, but with impedance mismatches you will see them most at the quarter wave. Two quarter waves (half wave) is optimal, but three again is worst. See this post on the matter:

https://lowpowerlab.com/forum/rf-range-antennas-rfm69-library/connecting-sma-pigtail-to-moteino/msg15744/#msg15744

The curved lines around the antenna elements and feedline is a visual representation of the current distribution. Alternating Current in an antenna is what makes it radiate. More current (line further away from the element) equals more radiation.

In the first example you'll see high current at the antenna feed point and almost no current on the feedline. Second picture is worst case scenario. A lot of current is present at the feed point on the outside of the coax. From Ohms law we know that when there's high current, there's a low voltage. Now look at the bottom of the quarter wavelength long feedline. The current line is pretty much 'touching' the feedline to indicate almost no current. This means high voltage. That can kill your radio.

My plan is to write state machines that will transmit state changes to one another.

That sounds like a fun project. Please, do share when there's something to show!

[sid1202]

Oh I see. you want the reflection to be at the same point as the transmission and not at the opposite range.  Basically the difference in potential between the signal and the reflection creates voltage and the voltage can kill the radio.

And I thought (how hard can using FM be ?) LOL turns out there are a lot of beasts all intertwined to make FM work right. This is quite the learning experience.

That sounds like a fun project. Please, do share when there's something to show!

I have been taking pictures. I'll be taking some video's soon. My first step is to change the throttle control from manual to electric so I can change the engine speed via remote. I'm in the middle of that at the moment. I will post the project once I'm done.

[captcha]

I find it fun to post some of my projects (here or elsewhere) as a sort of memento as well. Sure, I save pictures locally and have all sorts of datasheets to go with it, but posting something on a forum forces me to be to the point, concise and.. have a little fun with it. See how you go. We'll be here..

turns out there are a lot of beasts all intertwined to make FM work right.

With most of my answers I actually had HF SSB in mind. That said, what holds true on HF mostly applies to UHF as well.

This is quite the learning experience.

Mission accomplished, haha..

[d00m178]

Hello,

I have the same situation as topic starter - I need use external antenna for Moteino with LoRa 915Mhz

I've read all the information in this thread, and others related to this question and this is what I have:

- my RF cable is Radiolab RG-58 A/U (1x1.0)
impedance is 50 Ohm
vF is about  1,28 - I also found another values for vF, for example here: http://www.febo.com/reference/cable_data.html says that vF is 73.0
another sources says about 0.66...
I decided to use 1.28.

Using the calculator here:  http://n1gnn.com/webData/CoaxVelocityFactorLength.html
it says:

For coax with Vf = 1.28, 1/2 wavelength = 20.98 cm, 0.688 ft & 8.26 in

so to find length of feedline I need multiple 20.98?

for example 20.98*18 == 377.64 cm will be OK as feedline from Moteino UFL connector to my DIY GP antenna (like here: https://www.thethingsnetwork.org/forum/t/very-simple-homemade-outdoor-868mhz-antenna-groundplane/3160)?


but I use 1/4 wave length wires in my GP antenna (82mm for 915 MHz) -  can I use in this case 1/2 wavelength value (20.98) for calculating total length of feedline?
I see that 1/4 value (10.49) give me the same length 377.64 cm.


Please confirm or not my calculations.