Antenna Length and VSWR effect (reflected power from antenna)

[sparky]

First I would like to say that I'm very new to all this so forgive me for the noobie questions moving forward.

I purchased a couple RFM69HW 868/915 MHZ Moteino's and I'm just about finished soldering everything up.  I was about ready to solder the antenna's but had a question about the length.  According to the Moteino "pinout" page, the length should be 86mm (868) / 82mm (915).  The antennas that came with mine are both 79mm. 

After reading some posts on damaging the units from either incorrect antenna or wrong settings I wanted to make sure the lengths are correct.

Thanks,
GM 

[perky]

The lengths quoted are for 1/4 wave wire type antenna, however you can twist or coil the antenna to add inductance and shorten it's length and it will still be 1/4 wave. Does this 79mm antenna have a spiral around the edge?

Mark.

[sparky]

No, it's just a straight piece of wire that was shipped with the moteino.

Thanks

[perky]

Might not make a significant difference. There's a theory for dipoles that suggests slightly less than the calculated length results in resonance and practically the same radiation pattern, probably also true for 1/4 wave. There'a also not much of a ground plane either so I'm tempted to say this slight shortening could even be beneficial:
http://www.antenna-theory.com/antennas/halfwave.php
Mark.

[sparky]

So being these are monopole, would the same be true?

[perky]

I don't know the answer to that :-(

[Felix]

Basically the shorter length is to compensate for the PCB traces that lead to the actual matching network on the radio modules.
Either way the manual cutting of these is not an extremely precise process and in practice small differences don't seem to make a difference.

[sparky]

Basically the shorter length is to compensate for the PCB traces that lead to the actual matching network on the radio modules.
Either way the manual cutting of these is not an extremely precise process and in practice small differences don't seem to make a difference.

Ok, just didn't want to overheat anything.

Thanks

[captcha]

Hi Felix,

With regards to 'compensate for the PCB traces', I'm curious.. Has anyone ever published any useful data on that?

[Felix]

I don't think so. Just common sense and practical feel while not drifting too far from the theory. The one way you'd test this is with a VNA.

[captcha]

I know from experience when tuning my 433Mhz quarter-wave vertical and getting close to resonance I was cutting off bits 1mm at a time to get to the lowest SWR. Things are twice as sensitive on 915MHz. I wouldn't exactly say that it isn't a precise process. Will things work? Probably yes, but you're way off the scientific method.. ;-)

[perky]

Given there's little ground plane and is pretty sensitive to orientation and other surrounding things, are those few mm really going to make a big difference? I don't think so, I think there are so many other factors that would have an effect. I've tried manually snipping bits of antenna before and quite frankly moving the unit a little throws it right out again.
Mark.

[Felix]

I know from experience when tuning my 433Mhz quarter-wave vertical and getting close to resonance I was cutting off bits 1mm at a time to get to the lowest SWR. Things are twice as sensitive on 915MHz. I wouldn't exactly say that it isn't a precise process. Will things work? Probably yes, but you're way off the scientific method.. ;-)

I agree and I think I've done that long long ago when I was trying to get an accurate feel for the antenna length that I will ship to thousands of folks. I've arrived at the given length and it has never really been an issue and folks report ranges of hundreds of meters in the open, or even 1km+ with these wires by tweaking the settings. I'm sure things can be improved on a case by case basis but sometimes people bend the antennas and enclose them in cases or wire them in ways that alter the antenna performance way more than a perfect length. Plus the ideal performance is achieved when the antenna is perfectly perpendicular to it's "launch-pad" - the GND plane - which is almost never the case, even in my own projects.

[captcha]

Real-world scenarios:

but sometimes people bend the antennas and enclose them in cases or wire them in ways that alter the antenna performance way more than a perfect length

Quite true.. Fortunately, the RFM69HW is tolerant of up to 3:1 SWR* (a reflection of 25%). Pretty amazing unit..

Still, I wouldn't mind having access to that extra 25% of output power. Especially when the module only outputs 100mW.

* Reference: Page 11 of the datasheet: http://www.hoperf.com/upload/rf/RFM69HW-V1.3.pdf

[Felix]

In practice, sometimes the antenna has to loop around inside a box, or around the enclosure or even around the PCB itself. I expect that to cause gross VSWR (coefficient power reflected back from antenna). Even the projects where I've done this have worked without a problem although with a significant loss in RSSI at the other end.

Whenever possible keeping the antenna perpendicular to GND will boost the RSSI, and allow RFM69_ATC to reduce power and hence save battery on low power nodes.
It's all a tradeoff.

Here's a good read on VSWR: http://www.antenna-theory.com/definitions/vswr.php