tc replied the topic: 3dB Hybrids

The quadrature property of the coupled lines is a subset of the amazing properties of lossless, symmetric four-port microwave circuits. For double symmetry circuits (like two coupled lines or a hybrid coupler) the 16 S-parameters of the scattering matrix of the four-port reduce to 4 independent S-parameters: S11, S12, S13, S14.

Because the circuit is lossless (no heat or radiation losses), conservation of power is applicable to the four ports. This means that the multiplication of the scattering matrix by its complex conjugate transposed matrix will be equal to the unity matrix. This results in 16 equations - one for each element of the unity matrix. 12 of these equations are not independent and are just variations of the first four. So the 16 equations reduce to four complex equations involving the four complex S-parameters: S11, S12, S13, S14.

Go to www.microwaves101.com/encyclopedias/1158...metric-coupled-lines if you would like to see the
detailed four simplified equations.

If you simply by assuming S11=0 (and by symmetry S44=0) and assume S12 and S13 are non-zero, then the phase difference between
S12 and S13 is pi/2 exactly.

All the lossless symmetric four-ports in the world must satisfy these equations. Let’s just look at one of those circuits – a two line coupler. From the even and odd mode analysis provided by Microwaves101.com, if the values of Zoo and Zoe satisfy the equation Z0*Z0=Zoo*Zoe , then S11=0 (and by symmetry S44=0) and the four equations reduce to two equations.

If you assume that S12 and S13 are non-zero, then the phase difference between S12 and S13 is pi/2 exactly.

This is true for all frequencies, for any value of coupling and for any length of the coupled lines! If you don’t believe this, then plug into AWR Microwave Office and it will compute that S11 and S44 have a return loss and isolation greater that 300dB over the frequency range from 1 to 200 GHz Then the phase difference between S12 and S13 is pi/2 exactly regardless of the length of the coupled lines. For more information on this approach to the properties of a lossless 4-port, check out the textbook by Mongia, Bahl & Bhartia, RF and Microwave Coupled-Line Circuits, Artech House, 1999, pp 40-46.

madengr replied the topic: 3dB Hybrids

Look into the I/Q modulation; i.e. gain and phase imbalance, and isolation. The math should be the same. I was doing a hybrid fed, orthogonal feed patch for circular polarization. The isolation between LHC and RHC uses the same charts for gain & phase imbalance for image suppression in direct conversion. Though I haven't seen one for I to Q isolation, but the DSP guys have this all figured out for their corrections. I was able to get ~30 dB suppression which was good enough for my application, but some of these radio systems need 60 dB. I thought I'd be smart and put a PA on each arm of the hybrid, and of course they were run into compression, so AM-AM and AM-PM modulate the antenna polarization. Ended up having to match gain & phase on VNA when driven into compression, and group pairs by lot codes. The AM-AM and AM-PM varied drastically between lot codes; sometime compression, sometimes expansion. Real hair puller; lesson learned.

www.rfcafe.com/references/electrical/quad-mod.htm

Desert Sage replied the topic: 3dB Hybrids

Oh thank god for an answer. Thans MAD.

I guess I am asking about a real, physical device and not an abstract perfect hybrid. When I look at specs, I see 25 db isolation. This means for 3db coupler directivity of 22db. This isolation comes from leakage between arms, independent of load mismatch. It would be there with perfect load. Physically, you would want coupling only in the coupling region and zero coupling outside that region. This is practically impossible.

I will get back to this when I have time and report findings. I guess ultimately I am interested in how isolation hurts quadrature.

madengr replied the topic: 3dB Hybrids

In a perfect hybrid, the isn't the isolation (i.e. the return loss at input port) a function of mismatch between between the two loads? So at least for the phase and impedance mismatch between the output lines, those would be equivalent to a load mismatch.

Desert Sage replied the topic: 0dB Attenuators

Yes, series short circuits and shunt open circuits is a reasonably good 0dB attenuator. Depending on the frequency you may want the short to be a 50 ohm line.

Rykk replied the topic: 0dB Attenuators

Ok, so what exactly is in a 0dB pad? I'm trying to put one in front of an etched planar filter but I think I need some sort of actual path to ground. Is a 0 dB pad just a pi configuration with a short in series and opens for the shunts?
Thanks, Rick

Microwaves101 replied the topic: 0dB Attenuators

I like the other answers to this question, but let me add....

When designing a transmit or receive chain, it is always best practice to put in a little more gain than needed, and configure chip attenuators of low value (2 dB for example) wherever they have the least hit on performance (in front of a mixer, not in front of an LNA). Then when you build a prototype you can tweak the performance by swapping attenuation values in and out (so-called select-at-test components). In an example case, your customer comes by and says that you need to reduce DC current to meet a new requirement. You reduce the drain current in a driver amplifier, and its gain drops. Time to reduce that attenuator... you will be glad that you have a zero dB value available.

Good engineering is all about leaving options in a design that prevent an entire design re-spin. Software people call these options "hooks", they stole that word from us. Hmm, its about time to add that to the slang dictionary...

www.microwaves101.com/encyclopedias/microwave-slang#hook


Steve

madengr replied the topic: 0dB Attenuators

I use the SMT ones all the time. As mentioned above; phase. Recently used them to balance gain on a microstrip quadrature hybrid feeding two amps. The 0 dB attenuator will give a better phase match to it's companion attenuator than substituting a jumper or lead frame. It will also have similar parasitics and discontinuities, which is important at high frequencies. I wouldn't even buy a series of attenuators unless they offer a 0 dB or 1 dB value. I use these a lot:

ims-resistors.com/wp-content/uploads/2016/05/A-series.pdf

Desert Sage replied the topic: 0dB Attenuators

Suppose you are building many of some subsystem, like a converter for example. There is likely to be variation in the gain stages and the designer decides some of these variations can be reduced by using attenuators. The highest gains are reduced to the lowest gains. Now suppose there is cabling involved. For the lowest gain devices where no attenuation is required, if it is removed, then the cable may be too short and might need to be changed to a different cable. Think of the 0dB attenuator as a place saver so that only one cable is ever needed.

That's the first thought that comes into my head. Maybe somebody else can come up with other reason (phase?)?

Microwaves101 replied the topic: MW Power divider (WPD, GPD), Bandwidth extension techniques

For a broadband Gysel, Ali Darwish published a paper that shows 100% bandwidth. The design uses coupled lines. Here is the reference, contact me if you need a copy...

ieeexplore.ieee.org/document/7540294/

I plan to post a page on that combiner in the coming months.

Steve

mattskee replied the topic: MW Power divider (WPD, GPD), Bandwidth extension techniques

The best reference I personally have for the principles of multi-section or tapered line transformers is Pozar "Microwave Engineering", in the chapter on matching networks. Or there are probably some things online, maybe on this site: look up things like the Cheybshev transformer or Klopfenstein taper.

Microwaves101 replied the topic: MW Power divider (WPD, GPD), Bandwidth extension techniques

Hello

The bandwidth extension of a WPD is related to impedance transformers.. in a fifty ohm system you are transforming 50 ohms to 25 ohms (two 50 ohm terminations in parallel). For a simple WPD, the arms are 70.7 ohms and act as the entire transformer. Adding a second section either to the input, the outputs or internally means the transformation has additional elements. More elements means more BW. The individual transformers are quarter-wave at center frequency, but if you have two of them you will see that the length becomes quarterwave at F0/2, so you get more bandwidth below the band.

Transformer theory was well known before Wilkinson came along. You can go down a lot of different paths, like equal-ripple, max-flat, etc. There is no singular correct way to do this.

Sorry this is not a great explanation, maybe someone else can expand on this...

Steve

Desert Sage replied the topic: stripline(buried) filter??

A well built microstrip filter doesn't radiate badly. Stay away from parallel coupled line filters though. They will (launch surface modes, radiate, etc.).

Dave replied the topic: stripline(buried) filter??

If you've got access to a decent CAD tool (Sonnet, CST, MWO/Axiem, HFSS, etc) you can check that the structure won't exite unwanted modes.

The bigger challenge with stripline filters is that you can't easily get in and fiddle with line lengths/impedances to tweak the response. But depending what you're after (frequency, IL, VSWR, skirts, etc) again a good CAD tool will help you get pretty close. Sonnet have a free version which IIRC is limited only in model size.

Tomassoni replied the topic: stripline(buried) filter??

Yes, it is possible to obtain buried filters by using stripline technology. However there is a risk of exciting spurious modes, such a the spurious TEM mode and the parallel plate modes. To avoid the excitation of such modes, vias surrounding the filter can be exploited. The vias connect the top metal plate to the bottom metal plane.

Microwaves101 replied the topic: Coplanar Waveguide Marchand Balun

Group delay of anything that provides a flat phase shift in an interesting topic. Sometimes it adds confusion as it does not appear to make sense. In a rat-race used as a 180 divider, you will see a clear difference in the outputs, the 180 leg is always longer at center frequency. But the two delays intersect at band edges, and the shorter leg has more delay than the longer leg outside the band sweet spot!

In general, I would not put a lot of worry into analyzing group delay of a balun structure. But I will try to look at this, we are going to make a Marchand Balun page shortly.

Regards
Steve

macchiar replied the topic: Multiband power combiner with inputs of unequal magnitude and phase

Dear Khalid,
I don't think that the Wilkinson combiner is the best option. In fact, when the signals to be combined are unequal, some of the input signals power is dissipated in the resistors and the output (combined) power is less than the sum of the input signals power. Moreover isolation is relatively low.
In your case the best option is a selective combiner, but this requires that the signals to be combined are spectrally separated. In this case the combiner can be realized with a classical multiplexer (several filters with the output ports connected to a common junction). Using selective filters with high Q resonators you can obtain both high isolation and low loss in each channel.

Microwaves101 replied the topic: RF trace simulation in ADS

A long line that is mismatched will have a periodic response in S11, lots of dips and valleys. If you just look at one frequency you will have no idea what is going on. If you plot the trace across frequency and look at the maxima (-6 dB or whatever), then subtract 6 dB from that number (-12 dB in my example) that is the return loss that matters. The dips are caused by constructive interference due to the two mismatches, one at each end.

I hope I didn't misunderstand the question...

Steve

Giftina replied the topic: RF trace simulation in ADS

Thank you Dave for the link. The dielectric constant is 3.1 and the length of the trace is 42 mm. The maximum via seperation that the board fabricator can provide is 0.4mm.

Dave replied the topic: RF trace simulation in ADS

I don't know of any specific formulas for the distance from pin to first via. There might be some - check out Coplanar Waveguides and Circuits by Rainee N Simons. I haven't read it, but it's on the MW101 recommended books page: www.microwaves101.com/encyclopedias/book...icrowave-engineering I suspect in reality most people just treat it as an EM optimisation problem.

In the screenshot you posted, what's the dielectric constant and what's the line length? 77 GHz in free space is 3.9 mm, shorter once you take into account dielectric constant and filling factor, and I think you'll want the ground vias to be closer than that. Lambda/10 is a good rule of thumb at lower frequencies but almost impossible to fabricate at mm-wave.