Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] Impedance matching question

Status
Not open for further replies.

nenea dani

Junior Member level 1
Joined
Sep 5, 2009
Messages
17
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
Romania
Activity points
1,452
If I have two separate MMIC's GVA-13+ circuits at 10Ghz with a 20cm long RG142 cable and SMA sockets between them I have to connect a 50 ohm resistor to the ends (near every SMA)? Is there any relevance that there is a 4-5cm microstrip line (50ohm) between the MMIC output/input and the SMA socket? In this last case, I understand that there are 3 serial transmission lines or 5 with SMA, each with a characteristic Z0=50Ohm. Is it okay or I'm mistaken? I just do some experiments and I want to understand when to match and when not.
Thank you for your attention.
 

With ideal 50 Ohm matched amplifiers and reflection-free connectors, neither length nor number of cable sections would matter. The real MMIC reflection loss at 10 GHz isn't much better than 6 dB, you surely get some gain ripple caused by the cables. You want 50 ohm matching for the source and load, no resistors usually needed.
 
Thanks and I have another question. I have a lot of SMA sockets some cheap and others manufactured for 12 Ghz. Unfortunately, they were mixed in a box and I want to find the good ones. Is there a simple method or what device that would solve the problem? At least the cheap ones seem well machined and the dielectric does not melt in contact with the soldering iron, so I'm not easy to distinguish.
 

Not sure that the cheap connectors are actually worse. They just have a looser guaranteed specification.

I guess most reflections will be caused by the SMA to microstrip transitions, particularly if vertical sockets are involved. You need a 10 GHz capable VNA to monitor it. Or use FEM modelling to design the board.

To check the SMA parts without soldering it to PCB, you need a special text fixture in addition to the VNA. Its design is really advanced RF art. A reason why the "12 GHz" parts are more expensive.
 

Unfortunately, I have no device over 4.4 GHz. Under this frequency, all SMAs seem to have similar attenuation.
 

Is there any relevance that there is a 4-5cm microstrip line (50ohm) between the MMIC output/input and the SMA socket?
When is about the circuit noise figure, it will make a big difference where this relative long microstrip line is placed. If is placed to the input of the MMIC the noise figure of the circuit it will increase one-to-one with the microstrip line attenuation.
So for this reason is preferable to place the microstrip line to the MMIC output, where its influence to the circuit noise figure is lower.
 

I understand what I need to mount the microstrip line after the amplifier. I just jumped from HF to SHF and things are more than strange. It is unclear to me if I can display a link to a page of a ham to ask some questions about a stabilized oscillator at 10Ghz. I built this oscillator and I'm having trouble.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top