Tag Archives: Baluns
What is the difference between the BAL line of products (BAL-0003/6/9SMG and BAL-003/6/10) and the equivalent BALH products (BALH-0003/6/9SMG and BALH-0003/6/10)? Which one is a 1:1 (50 Ω single ended to 50 Ω differential/25 Ω single ended) transformer and which one is a 1:2 (50 Ω single ended to 100 Ω differential/50 Ω single ended) transformer?
A while ago we wrote a product feature for Microwave Journal for our isolation baluns. Basically we showed why they were better than 4 port VNAs for high speed differential testing.
These baluns are now available up to 67 GHz:
in the literature on the internet about baluns, a distinction that will often come up is ‘current’ baluns vs. ‘voltage’ baluns. I’ve always found this distinction confusing, because an electromagnetic wave converted by a balun from differential to unbalanced modes consists of both current AND voltage waves. They interplay to make a single electromagnetic wave. I can understand the terms applied to transistors, i.e. the current driven BJT vs. the voltage driven FET, since in this case the charges are either transmitted across the junction (current) or they only create a field (voltage).
If you really want to learn a lot about baluns, you need to talk to amateur radio operators. Jerry Sevick was probably more obsessed with baluns than anyone else in history. This means that the modern understanding of baluns is as viewed through the lens of antenna operation. The roots of the terms voltage balun and current balun trace to an article by Roy W. Lewallen, call sign IVTEL, in the ARRL Antenna Compendium, Baluns: What They Do And How They Do lt. In it, Lewallen describes the following two baluns, and differentiates one as the ‘current’ balun, and one as the ‘voltage’ balun:
This paper described the 1:1 transmission line transformer on top, the basis of our surface mount and 10 GHz and below connectorized baluns, as a current balun. This is because they force the currents to be equal and opposite, regardless of the differential impedances. This is easy to understand in an ideal flux coupled balun, as the signal transfers through the magnetic flux from one wire to the other wire only due to the current induction.
The second balun is Fig. 2 from Ruthroff’s seminal and diminutive named paper, “Some Broad-Band Transformers”. The connoisseur will recognize this as a compensated version of the above 1:1 balun. Lewallen describes this as a voltage balun, since the extra line forces the voltages at the output to be equal and opposite, regardless of what the impedances are.
I believe there are two problems with Lewallen’s analysis, one that is sort of his fault and one that is not. First, Lewallen points out in his paper that the issues arise with the different types only when the impedances are unmatched between the differential ports, as one would see with an antenna but not in many of the modern high speed differential setups that Marki baluns are suited for.
Second, it is based on a DC analysis, assuming that the voltages are not time varying. This is true at the low frequencies Lewallen was working at, but it breaks down for transmission line transformers generally. Also the impedance mismatches can cause reflections at higher frequencies that break down the ‘current’ balun argument.
In conclusion, I think that these distinctions may be valid and meaningful for low frequency baluns used in amateur radio setups, but for high speed test baluns the most useful metric is not voltage vs. current, but how much isolation is there?
As detailed in this blog post, there was a previously unaddressed problem with using baluns back to back in a test setup, with a VNA for example. The problem was that the baluns did not have isolation, which would cause a signal input to one differential port to show up at the other differential port. This would cause a resonance in the S parameters, and an ‘echo’ in the time domain behavior of the balun.
Starting now, this problem is a thing of the past. Marki Microwave just released a new line of ‘Isolation Baluns’. These are baluns built using a through line and an inverter, as has been sold before, but with a Wilkinson power divider between them. As I showed in this blog post, a Wilkinson power divider is capable of splitting data. It is also capable of combining data, so long as the data is common mode (identical on each arm). This is the circumstance with the new baluns, when placed back to back.
Previously this would cause an insertion loss ripple, like this:
As you can see in this second plot taken with back to back baluns there is very little ripple in the insertion loss, vs. the large ripple in the insertion loss of the back to back baluns in the case with no isolation. Additionally, the return loss does not exhibit the same high values, for the same reason.
This effect can be seen in the time domain as well. When two non-isolation baluns are placed back to back, a step echo can be seen in the oscilloscope trace of a square wave input:
The top trace here is the input, and the bottom trace is the output of two back to back baluns without isolation (not to scale). Compare this with the effect when two isolation baluns are placed back to back:
This implies that these isolation baluns can be used to extend a 2 port VNA to differential testing simply by de-embedding the S-parameters of the back to back baluns, something that was not possible before. This is something that we are still working on, but I will write an update when we have an exact procedure.
In making the datasheets for the first Microlithic frequency doubler (MLD-1640), it occurred to us that not enough has been made about the difference between isolation and suppression.
In mixers and amplifiers, some parameters are expressed relative to the input powers, while some are expressed in terms of the output power, with the conversion loss or gain calibrated out. This includes third order intercept point (IP3), which can be expressed as either input IP3 (IIP3), or output IP3 (OIP3). In general it is better to use OIP3 for mixers, since what really affects the dynamic range of a system is the amplitude difference between the output signal and output spur, expressed in dB relative to the output signal or carrier (dBc). This is illustrated in the table below, where the difference between the T3 and competing mixers is even greater when the superior conversion loss of the T3 is considered.
|T3-05||33 dBm||6.5 dB||26.5 dBm|
|Imitator 1||25 dBm||10.7 dB||14.3 dBm|
|Imitator 2||30 dBm||9 dB||21 dBm|
Note that it is better to use IIP3 in amps, for the opposite reason, namely that you want to give the amp credit for it’s gain. So in parts with gain the appropriate measure is IIP3, while in parts with a loss the appropriate spec is OIP3.
When the same logic is applied to spurious products in mixers and multipliers, the input referred value in dB is called isolation, while the output referred value in dBc is called suppression. Suppression is the preferable number to use, because it expresses the important value to the system. The isolation can always be improved by increasing the conversion loss of the mixer or multiplier, but this is obviously undesirable. There are, however, some issues using suppression.
The first comes with mixers. In all mixers we express the spurious output of the LO in terms of isolation, since it is dependent on the input LO power. Since the LO power does not change the conversion loss referenced to the input, this means that the suppression can vary by several dB with different LO drive levels.
The second complication is that the input signal, converted signal, and spurious tone are all at different frequencies. For example, when using a doubler with an 8-20 GHz input range, the output doubled frequency is 16-40 GHz, and the undesired tripled frequency is 24-60 GHz. This means that the isolation curves look like this:
While the third harmonic suppression looks like this:
The curves look slightly different. The suppression is more stretched out, and distorted by the curve of the conversion loss of the doubler. This is the result of the suppression calculation, which can basically be thought of as a 1:1 mapping of the isolation through the conversion loss.
I got an email asking if our baluns would work to 16 Gb/s data signals, so I tested them with the fastest data signal we can generate here before our BERT starts to go loopy. Here are the results:
I was thinking about the difference between power dividers, baluns, and couplers, and realized that they could all be thought of as power splitters. The characteristics that make them different are the relationship between the outputs in terms of amplitude, phase, and attenuation between outputs. Here is a brief chart that explains them all:
As I’m detailing in an upcoming app note, baluns are extremely useful devices. They can be used to interface with differential chips, build balanced amps and mixers, and drive antennas. They can also be used to create differential signaling lines that are immune to common mode noise.
There is a problem, however, if the signaling lines are not sufficiently lossy enough. If two baluns are placed back to back, the insertion loss is about what is expected, but it has tremendous ripple in both insertion and return loss. The frequency of this ripple is determined by the length and amount of loss of the lines connecting the two baluns. Here is what the output of two BALH-0006 baluns back to back looks like:
As you can see the insertion loss ripples along the expected insertion loss curve, along with the return loss increasing dramatically at each insertion loss suckout. So there is obviously some resonant behavior going on, the question is where does it come from exactly?
The frequency of the insertion loss ripple depends on the length of the cables between the baluns, therefore there must be a stray signal rattling around on the cables. There are two potential sources for this signal: bad return loss on the output, and the lack of isolation. EIther the signal is reflecting from the differential ports of the second balun, or it is being created by the inputs traveling from one differential port to the other. Previously this signal has been blamed on the poor return loss, but this seems unlikely.
.The BALH-0006 has good output return loss, which is odd for a balun with no isolation. In general you would expect to see common mode signals (which is half of any input signal when the other port is grounded) reflected entirely. In the case of the BALH-0006, the return loss is actually better than 20 dB at low frequencies, and better than 15 dB across the band. Therefore it seems more likely that the cause is lack of isolation. The BALH has only 1 dB better isolation than it’s insertion loss, which means that almost as much power goes from one balanced arm to the other as from the balanced side to the unbalanced side. How do we prove that it is one or the other?
This is where time domain techniques become valuable. First define the delay time for each step in the balun connection. Through the balun is t1, through the cables is t2, and from one balanced input out to the other (through lack of isolation) is t3.
The through path, is delayed by 1 balun t1, then the cables t2, then the second balun t1, for a total delay time of 2*t1+t2.
A pulse generated by a bad return loss will see the first balun t1, then the cable delay t2, then it will reflect from the other balun’s balanced outputs and go back through the cables for another t2, then again reflect for another t2, and finally pass through the last balun t1. This gives a total delay of 2*t1 + 3*t2. We will see the step arrive at a time 2*t2 after the first step output.
If the ripple is dominated by the lack of isolation, the step will travel through the first balun t1, then the cable t2, then from one balanced output to the other t3, then through the cable t2, then from one balanced output to the other t3, then through the cable again t2, and finally out the last balun t1. The total time is 2*t1 + 3*t2 + 2*t3. The step will arrive 2*t2 + 2*t3 after the first step, and 2*t3 after the return loss step.
First we build a square wave generator with a low rep rate (200 MHz with a 20 ps risetime) by cascading several square wave amps. Next we measure each of the time delays. First the transit time of the balun is measured to be t1 = 330 ps (not needed for the experiment).
Then we look at the delay of the cable, measured to be t2 = 300 ps.
Finally the transit time from one balanced input to the other is found to be t3 = 100 ps. You can also see in the screenshot below that the output pulse from the second unbalanced input is still quite large.
Here is what the output of the two back to back baluns looks like, overlaid with the input.
You can clearly see that despite a clean input, there is a delay in achieving the full output power until some time after the initial step arrives, and further that there is some ripple on the step function in between. If we zoom in on the output pulse we can see the details.
Here we can clearly see that the initial step, and after a delay of about 600 ps (=2*t2) there is another very small step, and finally the big step 200 ps (=2*t3) after that.
So now we have the full story of what happens when you put two baluns back to back and try to send data through them. Even with a perfect return loss, the non-isolated path creates a delayed second version of the input step, distorting the output pattern. Eventually all of the power arrives at the end, but only after a long delay equal to the twice the cable delay (plus some small amount). This contradicts what is currently found in the applications literature, which blames this phenomenon on return loss problems.
When you master phase, you become like a God, capable of performing wonders that mere mortals can only dream of. Wonders like making laser beams (using phase engineered quarter-wave reflectors), communicate tremendous information great distances through thin air (all modern communication formats use both amplitude and phase), and create amazing products (balanced amplifiers, balanced mixers, phased array antennas, Mach Zender modulators, the list never ends).
BUT…phase is the hardest thing to understand in microwaves, RF, and photonics. It is hard to measure, hard to visualize, and makes some very confusing homework problems that kept me in the late night coffeeshops of Champaign-Urbana well past my bedtime.
In this post we will make a dent in the universe of phase understanding by clarifying the difference between phase and group delay, and in the process explain why you can’t match phase with variable line lengths. When you buy a phase shifter, it is sometimes what I would call a real phase shifter, and sometimes what I would refer to as a ‘group delay shifter’. The trombone type variable delay lines (we like the ones from sage) are actually variable time delay elements, and not phase shifters.
A group delay (or time) shift is easy to understand: it is how long the pulse (or wave) takes to arrive at your measuring receiver. Differential delay is therefore the difference in how long it takes for two pulses or waves to arrive. In passive components it is just the distance divided by the speed of light (or whatever your wave is) at your frequency in your material.
Phase is much more difficult. It is the integral of group delay over frequency (plus an offset), or differently the group delay is the derivative of the phase vs. frequency. This is why filters can be used as time delays; the edges of the filter have significant phase variation that leads to significant group delay variations over a narrow bandwidth (this is called Kramers-Kronig relation).
A variable length delay line, therefore, can only change the phase by changing the group delay. But by changing the group delay, you are changing the integral (slope) of the phase vs. frequency. This means that the phase change will be different at different frequencies. This is very different than what you get from a quadrature hybrid coupler, or a balun, where the phase shift is constant across frequencies. The difference is shown below. First is a plot of the phase difference between the two outputs of a BAL-0520 Balun (180°), a QH-0226 quad hybrid (90°), a coupler plus two 37.5° Schiffman phase shifters we developed as a custom (165°), a PD-0220 wilkinson power divider (0°), and a PD-0220 with an extra .570″ adapter on one side (variable).
As you can see, the phase is flat across the bandwidth of the device for everything except the PD-0220 with the extra delay line (adapter). This has a rapidly changing phase across frequencies. If we take the derivative of this we should get the group delay, but instead I measured the differential group delay with the PNA-X.
Here you can see that the differential group delay between outputs for each of the devices is 0, except for the power divider with the adapter, which has a flat constant group delay (ignore the big hump, I think that is from the calculation the PNA is doing with the phase flip).
So what is the lesson? You can phase match two outputs using a variable delay line, but only at a single frequency. Otherwise you have to do it with a coupler, a balun, a Schiffman, or some other true variable phase circuit.
Since we released our BAL-0006SMG (and now BALH-0006SMG) broadband surface mount baluns, we have received a lot of interest in them from people using high speed analog to digital converters, and have wondered why they were so driven to find the best balun/transformer available. Today I figured out why.
Rob Reeder, of Analog Devices has a paper on why phase balance in particular matters for A/D converters. The basic idea is that while an ideal ADC is perfectly linear, an actual ADC has a slightly non-linear transfer function, with some remnants of the second and third harmonics appearing in the output of the ADC. This is the limiting factor in the dynamic range of the A/D, and that is the reason that spur-free dynamic range (SFDR) is one of the most important specs for an ADC. The most significant source of signal distortion is frequently the second or third harmonic, and this is specified on the datasheet.
A differential ADC using a balun/transformer at the front end can cancel out the second harmonic significantly, but only if it is well balanced. Any imbalance, and in particular the phase imbalance, will lead to a significant increase in the second harmonic spur in the ADC output. The phase balance is more important than amplitude, because the distortion due to phase imbalance is proportional to the amplitude squared of the input signal, while the distortion due to amplitude imbalance is proportional to the difference in the square of the amplitudes, which is much lower.
So I took Reeder’s MATLAB code to compare the BAL-0006SMG to a competing 6 GHz balun. Here are the results. First the signal output with the competitor balun (amplitude balance .5 dB, phase balance 12 degrees at around 3 GHz), assuming an ADC that would be second harmonic limited:
In this case the dynamic range is limited by the second harmonic to 73 dB. Now if we replace it with a Marki BAL-0006SMG at the same frequency, the amplitude balance improves from .5 dB to .2 dB typically, and the phase improves from 12 degrees to 3 degrees typically. This is the result:
As you can see the spur is reduced to 85 dBc, near the third harmonic. In this case the change of the balun improved the dynamic range of the ADC by more than 10 dB. When high signal sensitivity in broadband applications is critical, the BAL-0006SMG is by far the best choice available on the market today.