IMPEDANCE MATCHING IN RF ON PCB BOARD USING HELICAL ANTENNA

[Nawsheen]

Hello everyone
I am working on this board in which I have to achieve 50 OHM impedance matching using the helical type of antenna and receiver as LIZARD IC but I am unable to get the desired value even after using the calculated values for inductance and capacitance,and also the range is not gained kindly help me in this.

 

[rjenkinsgb]

What is the physical mounting and electrical connection for the antenna, to the receiver?

The impedance of a monopole antenna is dependant on it being mounted on/over a ground plane or having some kind of counterpoise, radials or at least a metal case of some sort at its base.

Without something like that, there is no fixed length for the antenna itself to be resonant, so the connecting cable (or stripline etc.) becomes part of the antenna rather than being a feed line.

 

[Nawsheen]

If you are talking about the helical antenna than that is on the PCB board and the other antenna is connected physically through a connector that is working for the key fob and maintaining its range. The target is an antenna impedance matching of the helical type. To the receiver, the antenna is connected by making a network matching circuitry using impedances and capacitances..Below is the attachment u can check

 

[rjenkinsgb]

The target is antenna impedance matching of the helical type

But what is it mounted to?

A connector on a metal casing, or one in mid air, or on the end of a cable, or what?

A photograph would be helpful.

 

[Nawsheen]

It will be mounted inside the bike through the connector.

 

[rjenkinsgb]

You are still not giving the physical/mechanical details - they are absolutely critical for antenna impedance matching.

If the "lizard" you mention is the NXP device, then that will be working somewhere likely in the 400 - 900 MHz range.
At UHF it is easy to use a dipole (two helicals to be compact) or tuned loop, as things do not need to be particularly large.

eg. This is the receive antenna in a one of our products a few years back (a cropped section from the through-board view in our PCB software).

It is a tuned loop, impedance matched by a tap part way from the ground connection. Simple and low Q but very effective, working over a considerable range.

 

[Nawsheen]

OUR antenna is the pcb mounted helical type of antenna and the board is made for ECU module that will be inside black box and yes it is the nxp ic and our range is upto 433.92 Mhz

 

[rjenkinsgb]

You are still not giving the physical/mechanical details - they are absolutely critical for antenna impedance matching.

What is the physical / mechanical arrangement of connection, ground, and all other parts relating to the antenna.
You have still not given specifics.

Please provide exact detail or preferably a photograph, otherwise no one can help you.

 

[RadioRon]

Antennas are affected by the environment around them. This sensitivity to their surroundings affects their input impedance as well as their radiation performance. Nearby conductors and dielectrics both will affect the antenna, and the affects are stronger when these materials are closer rather than further away from the antenna. Some antennas can be designed to be less affected by surroundings, but the base fed helical antenna is particularly sensitive to conductors around it. This is because this kind of antenna requires a counterpoise, of which the “ground plane” is the usual kind. Another way of saying this is that the helical antenna part (the component that you solder down, or the pcb trace that you put on your board) is actually only one half of the entire radiating structure, the other half is the ground plane or counterpoise. RF currents that flow into the feed point of the helical must be accompanied by RF currents that flow into the counterpoise in the opposite direction. These counterpoise currents flow into the “cold” side of the transmission line that connects to the helical. For example, if a coaxial cable was used to connect to the helical, the RF current going to the helical antenna is that current flowing on the center conductor of the cable. That current is accompanied by an RF current on the inside of the coaxial shield that flows in the opposite direction and this shield current must have a place to go to satisfy a good impedance match where the cable ends at the antenna. That place is the ground plane, and must be a low loss conductor of some sort that is physically oriented to be away from the antenna element. The physical shape and orientation of the ground plane is critical and that is why a photograph of your pcb and helical together is the only way we can comment on its design affecting impedance match.

 

[Nawsheen]

Hello Everyone Thanks for the reply I am sharing with you the top and bottom layer picture of the pcb board you can check it and if this will help you then kindly reply

 

[rjenkinsgb]

It appears you are driving one end and the other is grounded via a capacitor (as in the diagram in post #3?)

A self-resonant antenna with one end grounded has theoretical infinite impedance at the other end.

It is just not a good way to design an antenna.
An antenna with a grounded end is matched by feeding a tap near that ground point, not the other end, which is the most critical point for resonance.

I do not think it can ever work well, especially in PCB form; I've never seen it used in any commercial product and they are usually optimised to the limit.

The usual PCB type are a simple end-fed type based on a quarter wave, whether straight or zig-zag, or a tuned loop with a tap at a suitable impedance point for matching, as ours above.
Or simple formed wire antennas, which have higher Q that simple PCB ones.

THIS is a commercial helical antenna for UHF modules - end fed with the other end open, a "compacted" quarter wave end fed antenna:

https://www.radiometrix.com/files/additional/coil-ant-right-angled.jpg

Edit - What exact RF IC are you using? NXP use the name Lizard in publicity material, but not in connection with IC data sheets...

 

[RadioRon]

How many layers of copper does your board have? If more than 2 layers, can you show the remaining layers that we have not seen yet?
I think that your antenna trace is probably long enough to provide adequate performance. It appears to be a zig-zag design but it would be helpful to know the dimension of the short end of the pcb so that I can determine how long the antenna trace is. It would be even more useful to have a scale drawing.

Can you please tell us what feed point impedance you are measuring and also describe exactly how you are coupling your network analyzer into the antenna feed point to measure its impedance?

I understand you are using NXP NCK2910 on 433.92 MHz. I do not have access to the datasheet for this part, so am limited in what I can suggest. It would be helpful to read the manufacturer's application information about antenna matching.

 

[rjenkinsgb]

I've found a link for the NCK2910 demo board manual, including schematics.
It does not open directly for me but works if saved then opened: **broken link removed**

See pages 12 - 15.

I notice that there is no DC path from the IC antenna pins in the makers schematic, just a harmonic filter & coupling cap, whilst the OP's version has an inductor directly to ground.

That could also be causing problems if there is supposed to be an internal DC bias voltage on the RF pins.

 

[RadioRon]

I did already view that powerpoint set so got a good idea of the functions of this IC, but it does not provide the application info I was hoping for. I agree with you that there is a possibility that a shunt L is not appropriate, but we are just speculating until we understand why the op has chosen the particular matching network that he has in his schematic. I'm wondering if the design of the zig-zag antenna is a from-scratch effort, or copied from an application guide or someone else's product. If the latter, then perhaps the op is simply not competent in impedance matching techniques and struggling a bit. Don't know until I get more info.

Assuming no groundplane under the zig-zag, I find that the conductor length is roughly 140 mm, but in the zig-zag configuration I would guess to shorten the electrical length to roughly 60% of that, or 84 mm. That would make this appear to be about 1/8 wavelength, and with a shorted end, which, if I'm not mistaken, supports your observation that input impedance will be relatively high but not as bad as if this was 1/4 wavelength. This is supported by the component values in his matching network, which appear to be supporting a higher-Z match.

 

[Nawsheen]

Hello Everyone,
Thanks for your reply. Regarding the number of copper planes that we have in our board its more than 2 layers, we have 4 layers out of which the topmost is the component then the ground plane then the power then the component again and the distance from the antenna is 5 mm also the layers are not playing any role. Which portion of the datasheet will be helpful for you I can share that.
And yes this is an ongoing project I can share the ABIC application note to you for your reference you can check it once.

 

[RadioRon]

I would like to see the table of electrical specifications and I would like to see all pages that discuss antennas, antenna matching, pcb layout guildelines for the antenna, and example IC and antenna layouts (for example, evaluation board layouts too). If necessary, you can PM these to me.