Big difference of AC analysis of LC tank: between simulation & Vector Network Analyser measurement.

Thread Starter

Ngô Đình Nhân

Joined Jan 8, 2017
45
Hello all
I got a problem with LC tank.
In particular, f0 of LC tank is ~ 5MHz (470uH and 2200pF)
When I did an AC analysis with TINA software (I swept from 3MHz - to 7MHz), gain at resonant frequency = 1, phase ~ 0 degree.
However, when I did an AC measurement with Vector Network Analyser ( I also swept from 3MHZ - 7MHz, and schematic circuit is same to simulator). In this test, I only got gain ~ 0.4.
I am so confused with this big difference.

Does anyone please help me to clear this problem ?

P/s: Please refer to my attached files.
 

Attachments

danadak

Joined Mar 10, 2018
4,057
Also are you loading tank with the 50 ohms of the VNA generator ? If so that
will de-Q the tank circuit severely. The BW will be very broad as a result of that
loading.

Regards, Dana.
 

Thread Starter

Ngô Đình Nhân

Joined Jan 8, 2017
45
Also are you loading tank with the 50 ohms of the VNA generator ? If so that
will de-Q the tank circuit severely. The BW will be very broad as a result of that
loading.

Regards, Dana.
Hello danada
I have a resistance (50Ohm) in series with LC tank. I mean TX(VNA) - R100Ohm - LC Tank - RX(VNA).
Could you please give me more clear about your opinion ?
Thanks
 

Attachments

MrAl

Joined Jun 17, 2014
11,389
Hi,

Yes parasitics are often overlooked in simulations and so you dont get the same as real world measurements. It takes a lot to get this right sometimes especially when dealing with low capacitances in the circuit already.
 

danadak

Joined Mar 10, 2018
4,057
When you did Tina sim you used 1K as R between tank and Vsrc. You should have done -

1) Use 50 ohms not 1K ohms to match what VNA is going to present to circuit

OR

2) Use a current source to drive Tank in TINA (present infinite ohms to tank,
effectively no load) and sweep the current source frequency.

When you use a VNA it loads the circuit attachment nodes with 50 ohms, so
either use a pad or a series R that is >> parasitic R in tank, coil.....so it does
not load the circuit under test appreciably.

Regards, Dana.
 
Top