315MHz or 433MHz External Antenna for 433MHz Receiver.

I have a functional project which uses 433MHz RF.
For the moment I am using a cheap receiver with a 433MHz spring coil antenna soldered directly to it.

I want to fit the project into an enclosure with an SMA bulkhead connector for an external antenna.
To do this, I would need to solder a piece of wire from the antenna pin/hole on the Rx module to the centre pin of the SMA bulkhead, which I imagine would count towards the antenna length.

1/4 wave antenna lengths are :
433MHz - 17.28 cm
315MHz - 23.81 cm

So my first question is - Should I use a 433MHZ SMA antenna OR make the connecting wire 6.5 cm long and use a 315MHz antenna?

Also, does it make a difference what kind of wire is used i.e. shielded coax or PVC insulated?

And, does the SMA bulkhead (outer pin) need to be connected to GND?

An out-there question - Does coax cable NOT count towards the length of an antenna?

FWIW I have a 433MHz whip antenna on order with 1.5m coax cable which has me thinking...

Should I use a 433MHZ SMA antenna OR make the connecting wire 6.5 cm long and use a 315MHz antenna?

Would you wear one brown shoe and one green one? Antennas and frequencies are made to be used together.

The distance between the connector in the housing and the radio should be relatively small (and ignorable).

aisc:
To do this, I would need to solder a piece of wire from the antenna pin/hole on the Rx module to the centre pin of the SMA bulkhead, which I imagine would count towards the antenna length.

No, it would not count toward the length, because it is (should be) coaxial cable.

So my first question is - Should I use a 433MHZ SMA antenna OR make the connecting wire 6.5 cm long and use a 315MHz antenna?

For what you're doing, it makes little difference for practical purposes. That is because the optimum antenna length varies according to installation location. Many such antennas actually cover a range of frequencies. But it makes sense to aim for the correct frequency.

Also, does it make a difference what kind of wire is used i.e. shielded coax or PVC insulated?

It should be shielded, and have 50 ohm characteristic impedance.

And, does the SMA bulkhead (outer pin) need to be connected to GND?

For a 1/4 wave whip, it should be connected to a metal ground plane (for example metal chassis) and the coax shield.

An out-there question - Does coax cable NOT count towards the length of an antenna?

Once again, no.

@PaulS : Not sure how u arrived at that comparison. If you read my post, I was accounting for the different frequencies by proposing to make up the difference in recommended antenna length between 315MHz and 433MHz with the connecting wire, thus theoretically painting both shoes red - or green if u choose.

@ Aarg : Thanks for the input. I plan on using an SMA/U.fl pigtail bulkhead which will allow me to put an IPX/U.fl socket on my custom RF module shield.

My enclosure is ABS plastic and larger than the PCB. I will therefore need a back-plate to mount the PCB.
At present I have a piece of hardboard (composite wood) I plan on using. My setup lacks a metal plane to connect one side of the U.fl to.

I see 2 options :
A. I could use an aluminium back-plate as which would give me a metal plane, but I am of the understanding with RF signals I should avoid metal surfaces.
B. Alternatively, I could use a non-metallic back-plate and add a separate metal plane. For example I could mount the bulkhead through a suitable piece of metal - similar to a bolt-washer-mounting surface-nut arrangement, where the washer represents the metal plane.

Would you recommend option A or option B or maybe an option C?
Would the metal plane need to be grounded?

Many SMA antennas are not 1/4 wave, so they actually don't depend on a ground plane to operate correctly. These are sleeve dipoles, or else a form of end fed dipole. If you rip apart an old wifi antenna, you will often see this.

Having said that, for the frequency in question, many short antennas would be a short element with a loading coil to match the impedance. These would perform better with a ground plane.

I'm having trouble visualizing your back-plate, separate metal plane, bulkhead etc. Perhaps you can provide a drawing?

I have attached a sketch and some photos.

Some comments.
The enclosure has a base and cover section of almost equal height.
The stand-offs are moulded into the base.
The SMA bulkhead will be fitted through the side of the base.
The back-plate (actually probably base-plate in this case) is hardboard.
FWIW I am considering not using a back/base-plate and fixing additional stand-offs directly to the bottom of the enclosure to match the Arduino mounting holes.

The optional metal plane is shown in the SMA bulkhead detail and can be sized to suit.

RFAntennaSketch2.pdf (68.6 KB)

Referring to RFAntennaSketch2.pdf, The SMA bulkhead to pixufl cable approach will work fine with the magnetic mount antenna shown. In that case, the ground plane is provided by whatever conductive surface you attach the magnet to (it will couple to it at RF frequencies). In the case of a "stub" antenna as shown in IMG_0353.JPG, the effective ground plane is the device ground. Since the device is quite large, that will balance the antenna current somewhat. However, due to the random dimensions, the standing waves will not be perfectly cancelled, and will be influenced strongly by device dimensions and placement. Hence your "optional metal plane" with the bulkhead is not a bad idea, but may not make a huge difference in practice.

Of the two antennas and connection methods, the magnetic mount type will have better and more constant performance than the stub. That is with or without the optional metal plane.

A good rule of thumb for the radius of the plate, is 1/4 wave from the bulkhead to the edge. If you can't do it in all directions, you could do it in segments. This type of RF counterpoise is called "radials". Again, the stub type is a compromise antenna, so my recommendation is to omit the metal plane. It will probably make very little difference.

No antenna will work as well as theory predicts, if ALL the connections are not properly matched in impedance.

Unless you know the impedance of the output pin on the transmitter board, you are just guessing at the other details of the antenna construction.

You may find that a simple 1/4 wave wire outperforms fancier antenna designs connected by an improperly matched cable.

jremington:
No antenna will work as well as theory predicts, if ALL the connections are not properly matched in impedance.

Unless you know the impedance of the output pin on the transmitter board, you are just guessing at the other details of the antenna construction.

You may find that a simple 1/4 wave wire outperforms fancier antenna designs connected by an improperly matched cable.

For practical reasons, my main objective is to have an external detachable antenna. Of course I would also like to have it work well.

What is an easy/good way to measure the impedance of the output pin?

aisc:
What is an easy/good way to measure the impedance of the output pin?

There is no easy way.

A good way is to use an RF network analyser but they are very expensive.

You will just have to assume that the output pin is about 50Ω.

Russell.