Return loss bridge testing

I want to make one return loss bridge. I found the following circuit. Is this ok?
I want to make it wideband upto more than 500 Mhz.
How to test it to find that it is good.

Hi,

The circuit you show in the drawing can be good to even 500MHz but in practice it needs much care in choosing/selecting the components, mainly the resistors and the wideband transformer.
For calibration, you have to use wideband terminations for 50 or 75 Ohms or 100 Ohms etc or you can take your RL bridge into a university microwave lab to compare the performances.
You have to check if your own terminations are real up to 500MHz. If you do not have any wideband terminations, you must search around your friends circles or nearby university labs etc. Without dependable terminations you cannot check or calibrate your own RL bridge. So start construction by keeping these in consideration.

See this link for a practical and sensitive RLB up to 500MHz:
http://www.qsl.net/n9zia/rlb/index.html This circuit eliminates the use of a wide band transformer.

There are some other links if you use google search engine but the above link is good for you I think.

unkarc

Hi

Please check https://www.edaboard.com/viewtopic.php?t=99369

You can get some information.

Thanks uncarc.
I hv gone to the webpage u hv directed.
In the schematic there is 50 ohm microstripline shown. I want to make the rlb for 75 ohms.
Guide me how will i make this 75 ohm microstriline?

Hi,

Go to this link on microstrip lines:

http://www.microwaves101.com/encyclo...microstrip.cfm

And you can find an on-line microstrip calculator to get the sizes for any impedance you need:

http://www.microwaves101.com/content/calculators.cfm

Of course you have to know the dielectric constant εr (also known as the relative permittivity) of the printed circuit board material you are going to use for the RL bridge.

If you tend to use FR-4 material, here are two links on its dielectric constant:

http://www.cis.syr.edu/~tksarkar/pdf/2001_Nov.pdf
http://www.emclab.umr.edu/pdf/TR00-1-041.pdf

unkarc

unkark,
u r perfectly predicted that iwill use FR4 material. Now from the another webpage i understand that for microstripline double sided PCB will be required.
Hence what to do if single sided pcb is used. How much distance between track & earth to keep for 75 ohms for normal design. It means that if the track width is 2mm then how much distance away the shield/earth should be.
In the project at QST i don't see any microstripline on the PCB. As it is 75 ohms connector will be there at the input & output in my project.
The wideband transformer in my rlb circuit decides the bandwidth & linearity. am i right?

Hi,

I suggest using your single sided PCB in two layers as a "sandwich" : on the upper copper side of the upper PCB you create the track and you can strongly attach the full copper side of the lower PCB to the bottom side of the upper PCB.
This may sound 'clumsy' but I cannot suggest a better solution for your having a single sided PCB... Of course you remove the unneccessary copper from the upper PCB only to have the track or tracks on it.
The distance between the track and the earth now works out to be thickness of your FR-4 single layer material and if you measure this and use in the on-line microstrip line calculator in the link I had included earlier, you will find out how much track width will be needed for your 75 Ohm example (use 4.6 for εr).
The reason you do not see microstrip line in the QST article is that it is needed from roughly about 150-200MHz and higher and they probably wanted first to show a measuring principle.
Yes, MAINLY the wideband transformer will determine your final bandwidth and linearity if you choose to build it.

unkarc

It took a more time to make the decision.
I will use double sided FR4 pcb because it will be easy to make then to make sandwitch of the two.
Now i will hv to design 75 ohm microstripline for this.
I will choose 35 micron or 70 micron 1.5 mm thick fr4 pcb. Help me to design microstripline for 75 ohms. This is just to confirm that my calculations for 75 ohm microstripline r correct.

I think relative dielectric constant of FR4 is 4.2. ( i think)
Then the track width for 75 ohm is 1.35mm. The height of the track is .035 mm = 35 microns.

Hi,

Yes your calculation seems very close for 75 Ohms with εr=4.2, double sided FR4, track-height=0.035mm, and here is some addition:
1) My friend's calculations in Serenade software gives 1.417mm for track-width when there is a metal cover plate 30mm above the microstrip
2) if you use a 0.010mm tin (Sn) layer on the FR4 microstrip copper, the track-width reduces to 1.407mm, (metal cover is also 30mm above the microstrip)
3) When you use less than 30mm distance between FR4 and metal cover/lid, the track-width reduces a little:
distance: track-width (untinned)
30mm : 1.417mm
15mm : 1.410mm
10mm : 1.398mm

regards
unkarc

Thank u unkarc,
This is more detailed for me.
Metal cover will be at the top. It should 30mm away from topside of pcb.
The bottom will be total ground. OK?

Yes, the bottom of the double sided FR4 will be total ground, this is the usual practice.
Thanks for the points.

unkarc

Thanks once more for quick positive reply!
http://www.qsl.net/n9zia/rlb/index.html
If i use two ad8307, one for gain & one for return loss, can i get both the graphs displayed on monitor?
This means, can i see gain & return loss at a time on a monitor? This is like u see on network analyzer.

Hi,

Well, if you wish to know the amplitude on the Unknown Port of the RLB, you may use another AD8307 for that purpose. Of course you have to design the PCB to accomodate the second log amp. If you study the data sheet, you will find application on how to use AD8307 as a single input log amp and not a differencial input (for instance Fig. 41 in http://www.analog.com/UploadedFiles/...ets/AD8307.pdf ).

Regarding the display of the second AD8307 you could use a simple analog meter instead of the digital one because it is much better to see amplitude changes up or down on a meter needle than on digital display. At least that is what I think...
Of course you may use the same display for the purpose like the one for the RLB if you design a switchable input for the display to select any one of the outputs.

unkarc

Thanks unkarc,
I think i am not able to write properly.
I want to get gain & return loss display on same oscilloscope/monitor at the same time. This way i can tune the filter of amplifer for frequency response & also see the return loss at a same time.
The following page have got project like that but it is software based usb interface.
http://n2pk.com/VNA/VNAarch.html

From this page u just see the display which i hv attached here.

I am going to use sweep generator as input & output i will see on oscilloscope/monitor.

Hello,

Now I understand, sorry but you seem to choose the hard way of achieving the goal. The hard way I mean to sense and process the swept amplitude info from the RL bridge and make it appear as an overlay on the same display where the return loss also appears. I do not have the proper knowledge to show you or guide you through, sorry for this. I think the link you included (very good) deliberately chose a software solution for the process, to simplify the hardware needs.
I wish you good luck in building your RL bridge and develop it further, perhaps other members here will be able to suggest suitable circuits.

unkarc

Similar analogue equipment hv been made by wavetek 1076 before 15 years. So when log detectors r available now so easily so i thought why not try to make one?

Return loss bridge testing :

Let me know how will i confirm the frequency bandwidth, range & accuracy of the return loss bridge.