MLX90130 antenna matching

[sukhmani]

Hello,

I had designed a balun and pi matchig network to match 50 ohm load impedance using IC MLX90130. I have got a new document and modified my circuit according to the values given, but that does not help. The problem is using smith chart values and LT Spice simluation, I am able to get desired output values, but after soldering components on PCB and lochraster, I dont seem to get the desired output.

Also another confusion is what is the exact impedance of MLX90130. Is it impedance or purely resistive? Data sheet gives various values like 34 ohm, 16 ohm, 27 ohm. And since MLX90130 had TX1 and TX2, the whole impedance should be twice of 34/16/27 ohm? And for calculation of balun values, should the DC blocking capacitors, EMI filter values be cosidered? And since balum does matching too, so technically there shouldnt be a need of a pi matching network following the balun!

MLX90130 provides 300mW of power and uses 5V supply. The voltage after balun should be 10V, (is that correct?) but LT Spice shows aound 6-8V depending on how balun is designed.

I would be grateful if someone could throw some light on this and I could mov further from this point. Right now I am going around in circles and confused.
Thank you

 

[FvM]

As most 13.56 MHz transceivers, the TX stage uses differential CMOS push-pull outputs. Assuming a pure resistive output impedance as specified in the datasheet should fit reality. If the design doesn't work in real hardware, I would expect inductor ESR as the most important cause for bad performance of your hardware, but you should be able to measure the exact network behaviour. The Melexis datasheet has no 50 ohm matching network, just the usual reader coil matching.

 

[sukhmani]

Like you said the melexis data sheet gives reader coil matching. So using a lattice balun with values L= 470nH and C= 390pF, matching to 50 ohm load should be achieved using smithchart simulation. But I will check for inductor ESR cause because the hardware is not giving the ouptut.

Thank you

 

[FvM]

A rough estimation gives ESR numbers of 1 -2 ohm at 13.56 MHz for suitable SMD inductors, not negligible but all in all acceptable. The networks shown in post #1 need better tuning however, the purpose of the picohenry inductor in the second circuit seems mysterious. I guess it's a typo. But to keep complexity within reasonable limits, you should better tune the basic circuit without pi-filter first.

Not too important, but the Rx input acts as a high ohmic impedance with a certain input capacitance. The AC voltage (fundamental wave )correspending to a 5Vpp square wave is about 3.2V rather than 5V.

 

[sukhmani]

Yeah, the pi network values are incorrect. I was just trying to manipulate to get 10 V output. The correct values according to smith chart simulation are c1= 730pF, c2 =232pF, L= 860nH.

The AC voltage, you mean at the input of the post #1 is a sqaure wave?

Is it possible that you know the valueof mlx90130 Rx value at 5V? Is it 34 or 27 or 16 ohm? That would also be helpful.

And I will tune the circuit without pi filter like you said.
Thank you