I have created an 5.8GHz patch antenna with complex impedance for my research (in CST MW Studio). But I have got a problem with the efficiencies (see picture). I think it's because the antenna is complex, but the port is not. I have already created a matching network (in schemetic) so the s-parameter seems to be good now. But how do I do the same with the efficiency so that it is matched for the complex antenna? Thank you in advance.
You are right. It is because your antenna has a complex impedance. Any part of the impedance that is imaginary causes reflection and not radiation. That is why antennas, and anything else that carry information or power need to be matched.
If your antenna is complex, then the driving circuit has to be complex, with the conjugate impedance, so that the imaginary parts cancel out.
You don't seem to have included your schematic.
The imaginary part of the port impedance needs to be the negative of the imaginary part in the antenna impedance. If one is inductive the other must be capacitive.
Thank you for your reply. Attached is the matching network and S_11 before and after the matching network. How can I use the matching network to get the correct values for the efficiency as for S_11?
The antenna seems to be correctly matched at 5.8 GHz. I do not know why the efficiency would be poor. With the good match you have there is almost no power returned to the generator. If the antenna is inefficient then power must be going somewhere. It may be going into losses in the microstrip. Is your substrate low loss?
I think the numbers in the figure you first included do not make sense if the network is matched properly. I think the radiation efficiency is the proportion of the power, that gets past the mismatch, that is radiated. -5 dB implies only 33% of the power is radiated, so 67% is probably lost as heat. The total efficiency includes return loss (S11) and includes the effect of power lost because it is reflected, so nearly -14 dB means about 9 dB loss due to mismatch as well as 5 dB lost in heat. 9 dB transmission loss is about right for 0.6 dB return loss, which was before you used the matching network, but not with the network. I think you have used the old wrong polar diagram figure to show the losses after you have applied the matching network. You should only have a very small difference between radiation efficiency and total efficiency, with that match.
You still have about 5 dB loss which might be due to your microstrip being too long or having a poor substrate. Have you used FR4, which is commonly used because it is cheap, but is not particularly good for any other reason?
Thank you very much again.
You said: "I think you have used the old wrong polar diagram figure to show the losses after you have applied the matching network." How can I use the correct diagram to show the losses with the matching network?
I am using Rogers RO4350B as substrate.
It looks like you showed us the polar diagram and efficiency data from before you matched the antenna.
The total efficiency will be different with the antenna matched. It looks like the total efficiency figure is from before you matched it, not after you matched it. You should have shown the polar diagram and efficiency data from after you matched it, not before.
There is still 3 or 4 dB to explain. Are your tracks long? How many dB/m does the Rogers microstrip calculator predict for the track you are using?
Yes you are right. But I do not know where to find the efficiency data after the matching. For example: For the S-Parameters I had to do a "Task" and then "Update" in Schemetic. But how do I get the efficiency after matching? Again "Task" and then? Maybe Transient? But I get an error for this.
"There is still 3 or 4 dB to explain. Are your tracks long? How many dB/m does the Rogers microstrip calculator predict for the track you are using?" Total loss is = 4.65378 dB/m (see picture). The tracks are approx. 20mm. So you mean -5db is correct, regardless of the matching?
The efficiency figures are shown in the bottom left-hand corner of the far-field plot. There should be a far-field plot for the matched patch antenna.
You can also see where the power has gone if you look under power in the results tree.
Why are you using 108 ohm microstrip feed? Your source will almost certainly be 50 ohms.
The 108 ohm line you have calculated for has loss of 4.65 ohms/m, and is 0.02 m long so has a loss less than 0.1 dB, so something else is losing nearly 5 dB. The patch will lose some as well, especially if it is narrow-band, because that increases the Q so it has high currents and voltages.
You won't improve this by matching, but you should be able to improve it if you find out where the lost power is going.
- Badges
- Science topic