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.

50 ohm matching in analog circuits

Status
Not open for further replies.

simha.nitc

Newbie level 4
Joined
Feb 13, 2013
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,326
Thanks in advance.

Can any tell me how to match analog circuits to 50 ohm line..

Lets say i have a differential buffer amplifier, and now i want to match the buffer to 50 ohm line. Effectively , gain=0dB, input impedance=100 ohm, output impedance=100 ohm and the Bandwidth is 100MHz.

Let me share the circuit too..
 

Attachments

  • image_preview.jpg
    image_preview.jpg
    12.8 KB · Views: 136

i am sorry ..i dint give you the complete details..its buffer amplifier which i am going to use in transmitter chain.. so its a balanced twin lead type only..
 

I will try to explain the way i understood..

Lets see the half circuitUntitled 1.png

Now to match the output to 50 ohm load..RL=RD=50 ohm

My confusion is ..if this is the way to match the load, we are loosing half of the signal at RD ,RL network ..

So, Can any one tell me..how to match analog circuits.. Here i can not use L or C ..
 

So, Can any one tell me..how to match analog circuits.. Here i can not use L or C ..
Misses the point. Analog circuits can be matched by LC circuits, but only for small bandwidth. Wide band mtching will use resistors and yes, loses part of the signal.
 

Agreed.. here i need 500MHz Bandwidth.. And one more difficulty is the values of L and C at these lower frequencies will be very huge which i can not use.. so i am looking for alternate way.. i have tried with resistor match..

But with resistor (RD=50 ,in this case) ..to maintain 0dBm gain (buffer), i need very huge Gm.. Becoz gain for the circuit is GmRD..

So i am searching for alternate matching methods..

- - - Updated - - -

I will try to talk in numbers rather than words.. Effectively i need 0dBm buffer which i have to match to 50 ohm load..
Now as RL=50, i need to keep RD=50.. Then the effective load is RD//RL whic is 25 Ohm..

Now i need Gain of 0dB, that is nothing but 1, GmR= 1

As R= 25 Ohm, Gm turns out to be 40m..

To achieve this much you have to burn so much of current..which is not desirable..

Thats why i am looking for the alternate way..


Please let me.. if i am missing something ..thank you
 

If the lower cut-off frequency is sufficiently high, RF transformers may be an option, otherwise multi-stage amplifiers with higher gain and pure resistive matching should be considered. If the frequency range includes DC, there's no alternative anyway.
 

Agreed.. But here i can not lose signal at DC .. so i am left with resistor matching..

If there is anything that i am missing.. please share it..

Thank you
 

Don't see any alternate to resistor matching.
You could perhaps add source followers to reduce the output impedance of the amp and more easily drive the 50 ohms.
 

Thanks in advance.

Can any tell me how to match analog circuits to 50 ohm line..

Lets say i have a differential buffer amplifier, and now i want to match the buffer to 50 ohm line. Effectively , gain=0dB, input impedance=100 ohm, output impedance=100 ohm and the Bandwidth is 100MHz.

Let me share the circuit too..

Normally source is matched to line unless it is a differential line then they are 1/2 each of line impedance.

here is an example of 100 Ohm differential line with 50 Ohm CML drivers with 3.2GHz BW and 325mV swing on Vcc=2.5V.

1883311600_1429033010.jpg
 
Thank you sunny ..

I want to discuss OP1dB of the same differential buffer amplifier posted above.. Can any explain how to relate the output swing to OP1dB?

or What can we do to improve OP1dB of the circuit ?
 

I think when Vce gets less than ~2V , compression starts to occur due to drop in hFE and compression rises rapidly <1V . when hFE drops towards 10% of peak value or so at saturation where Ic/Ib may range from 10 (typ) to 50 (Diodes Inc special parts I think are good choice. for low Vsat.) maybe.

Obviously choice of transistor makes a difference. I would choose low Vce_sat types from Diodes Inc for starters, but there may be many better options like the Current Mode Logic (CML) driver chip in previous example.

**broken link removed**
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top