Broadband Impedance Matching network

Hi Guys

I'm designing a square-law detector and I have to design a broadband matching network in range of frequency (800 MHz to 2 GHz) to match impedance source to the diodes. I'm using two Schottky diodes (HSMS 2852). The impedance source is 50 ohm and the load impedance of (two diodes) is (34.22-j314.311) in 1.4 GHz.
I need your help to design this matching network with this specifications.

Thanks in advance

Post your proposed circuit, maybe then we can help

My circuit is Square-law detector with voltage doubling (two diodes)

It won't be a simple network. What sort of return loss can you accept?

min S11 =-10dB

Is this just for a real application or just a theoretical exercise?

You may have to use multiple detectors and split the band in to a few sections.
You may then combine the output of the detectors.

If this is for a product design, then companies like AD make wide band detectors, which will be more cost effective in production.

This is real application....Can u please tell me ...How can I design the multiple detectors or how can I split the band into few sections?

Look at this. Works up to 2.5GHz
http://www.analog.com/static/importe...ets/AD8361.pdf

There is also an example to extend the range using a splitter.

Doing a match over such a wide band with discrete elements will result in something like a 7th order network, which will be difficult to replicate in manufacture due to component tolerances.

Thanks dear.........it is still this circuit complicated in this App. note bcz in my case I have to design three detectors to cover the frequency range.
I hope that u have another solutions.

regards

There is another trick we can use. We apply a slight bias (20-60uA)current to the detector diode. That may result in an easier match.

Yeah I knew before this solution but this will be effected directly to the sensitivity of detector and I need min. 8 mV/uW in this detector.

What is your lowest input threshold dBm or uW?

How broadband is 'broadband' ? You are not going to get a perfect match - there will always be some trade off between number of elements and the worst case VSWR which you can accept.

There is a spreadsheet here (second link down on page) which will help you get started with a simple matching network , either lumped or distributed:
http://www.rfshop.co.uk/blog/downloads

Remember that you'll need to transform the 50 ohm source impedance to be the complex conjugate of the diode impedance (34.22 + j314.311).

Check out the possible topologies for best BW using a circuit simulator, add more sections and reoptimise to improve the BW.

You can get better broadband matching when you use a 3dB coupler together with the detector diodes.

Plot below shows good return loss using this method, but with lower sensitivity of about 1.5mV/uW

Could u please post your proposed circuit?

Added after 7 minutes:

I'm designing this detector for a microwave radiometer
the lowest input power (modulated noise signal) is around -25 dBm

Added after 5 hours 9 minutes:

Could u please tell me, how did u design 3dB coupler?

regards

Added after 1 hours 31 minutes:

Dear E-design,

Thank you so much for your answers.....Could u please tell me, how did u design 3dB coupler?

Regards

The inductive and capacitive impedance's are about 1.414 times Zo. I started off with those values and fine tuned them as well as the termination for best return loss on the simulator.

http://www.odyseus.nildram.co.uk/RFM...les/Hybrid.pdf

http://www.microwaves101.com/encyclo...ers.cfm#lumped

you can read more

How much Zo is it? is it 50 ohm? What about the resistor R2 how did u calculate it ?....Is there 90 degree phase shift?

Thanks dear

Here diode characteristics for for the whole band is required to solve the impedance matching problem. You may try to use zmatch h t t p://w w w. zzmatch. com/ .

I extracted it and used it in my simulation. I include it here in Excel format

Sorry I didn't find the Excel format file.. could u please upload it?