An Inductively Assisted N:1 Switched-Capacitor Ladder Converter

summary

The gist

This paper proposes an inductively assisted N:1 switched-capacitor ladder buck (D-SCLB) converter that replaces the output-side switch in the bottom half-bridge of an N:1 SCLB with a small inductor…

In short

The D-SCLB converter is an N:1 switched-capacitor ladder buck converter that adds a small inductor to the bottom half-bridge during demagnetization. This modification lowers average inductor current to (1-M)Iout, allowing continuous regulation with M=D/(N-1+D). It improves output impedance by up to 30% and reduces average inductor current by up to 20% compared to standard SCLBs.

Key concepts

D-SCLB Topology
This is a modified N:1 switched-capacitor ladder buck converter. It introduces an inductor in parallel with the ladder capacitors specifically during the demagnetization phase of operation. This change helps manage current flow and enables better regulation compared to traditional SCLBs.
Average Inductor Current Reduction
The D-SCLB design manages inductor current during demagnetization, resulting in an average inductor current of (1-M)Iout. This is achieved by allowing flying capacitors to share the load current delivery, which is crucial for continuous regulation with M=D/(N-1+D).
Equivalent Output Impedance Model
The paper derives a closed-form model to predict the output impedance of the converter for any N. This model is valid regardless of whether optimal or uniform switch sizing is used, providing a predictable performance metric and showing significant improvements over SCLBs.
Switching Optimization (M)
M represents a parameter used in continuous regulation, defined as M=D/(N-1+D). By optimizing the inductor current management during demagnetization, the D-SCLB enables stable operation even when operating below the native discrete ratio of an N:1 SCLB.

Terminology used across episodes

This episode discusses

The paper

An Inductively Assisted N:1 Switched-Capacitor Ladder Converter · Read on arXiv

Sandeep Reddy Kukunuru, M Mahmudul Hasan Sajeeb, Rohit Reddy Karnaty, Loai G. Salem

University of California at Santa Barbara

This paper proposes an inductively assisted N:1 switched-capacitor ladder buck (D-SCLB) converter that replaces the output-side switch in the bottom half-bridge of an N:1 SCLB with a small inductor placed in parallel with the ladder capacitors during the demagnetization phase. The flying capacitors share the load current delivery, lowering the average inductor current to (1-M)Iout and enabling continuous regulation with M=D/(N-1+D). A closed-form output-impedance model, valid for arbitrary N under both optimal and uniform switch sizing, is derived and compared against SC, SCLB, FCMC, and DPSC topologies. For a 4:1 48V-to-PoL design, the D-SCLB reduces the average inductor current by up to 20% and improves the equivalent output impedance by up to 30% relative to the SCLB. Two hardware prototypes on a low-cost two-layer PCB achieve a peak efficiency of 96.2%, with the D-SCLB using 49% smaller inductor volume than the SCLB at comparable efficiency. Model predictions agree with measurements to within 1.5% under heavy load.

Transcript

Introduction to the show: ident: Robotics Radio. Generated commentary on the latest robotics and control papers.

Rosa: I'm Rosa.

Dev: And I'm Dev, here with Taro.

Rosa: We're looking at the paper: "An Inductively Assisted N".

Dev: Here's the gist of the paper. This paper proposes an inductively assisted N:1 switched-capacitor ladder buck (D-SCLB) converter that replaces the output-side switch in the bottom half-bridge of an N:1 SCLB with a small inductor placed in parallel with the ladder capacitors during the demagnetization phase because it lowers the average inductor current to (1−M)Iout and enables continuous regulation with M=D/(N−1+D).

Rosa: Converter Advantages The D-SCLB topology offers several performance improvements over existing topologies.

Dev: These advantages stem from how the inductor current is managed during the demagnetization phase, where the flying capacitors share the load current delivery, lowering the average inductor current to (1−M)Iout.

Rosa: Specifically, this enables continuous regulation with M=D/(N−1+D).

Title and authors: Rosa: So we're looking at this paper today, "An Inductively Assisted N:one Switched-Capacitor Ladder Converter," and the authors are Sandeep Reddy Kukunuru, M Mahmudul Hasan Sajeeb, Rohit Reddy Karnaty, and Loai G Salem <ref:2611.14778#pg1,Sandeep Reddy Kukunuru, M Mahmudul Hasan Sajeeb, Rohit Reddy Karnaty, and Loai>. They propose something called the D-SCLB converter.

Dev: That sounds like a lot of acronyms for a power converter design, Rosa. What's the core idea here? How does this D-SCLB actually help us with these N to one voltage conversions compared to what we usually see?

Rosa: Basically, they replace the output switch in the bottom half-bridge of an N to one switched capacitor ladder buck converter with a small inductor. They put this inductor right next to the ladder capacitors during a specific phase called demagnetization.

Dev: So, what's the main benefit they're getting from putting that inductor there? Is it just about making sure things run continuously?

Rosa: Yes, they say this setup lowers the average inductor current down to one minus M times the output current, or one minus M times I out. And because of that, it lets them maintain continuous regulation even when you're running a duty cycle of M equals D over N minus one plus D.

Taro: Continuous regulation is important for stability in real-world systems. But what happens when the system hits those boundary conditions where things aren't perfectly ideal? How does this design handle unexpected load changes?

Dev: That's exactly what I was thinking, Taro. If we can regulate continuously under those specific duty cycle constraints, it suggests better robustness during transitions in the load profile. Rosa, you mentioned a closed-form output-impedance model—what does that tell us about the performance envelope?

Rosa: Well, they derive a closed-form output impedance model that works for any N when you consider both optimal and uniform switch sizing. They compare it against other topologies like the standard SCLB, FCMC, and DPSC.

Taro: A model that works for arbitrary N is powerful because it means the performance metric isn't just tied to one specific conversion ratio; it gives us a general way to assess how well the design scales as we move up to higher N values.

Dev: That scaling aspect is interesting, but I want to talk about the practical gains. The paper claims for a four by one converter, which is forty-eight volts down to one volt, they see reductions in average inductor current and improvements in output impedance compared to the standard SCLB.

Rosa: For that four by one design, they found that this D-SCLB reduces the average inductor current by up to twenty percent and improves the equivalent output impedance by up to thirty percent relative to the standard SCLB.

Taro: Twenty percent reduction in inductor current is significant for power density and thermal management. That’s a lot of heat you don't have to worry about, right?

Title and authors: Dev: It is, but we also need to consider the hardware footprint. The paper shows they built these prototypes on a low-cost two-layer PCB and used a twenty-two microhenry inductor with a dc resistance of three hundred twenty milliohms.

Rosa: And what did they find regarding size? They compared the volume ratio of their D-SCLB to the SCLB, and it was zero point four nine, meaning the inductor volume is smaller. The total passive component volume ratio was zero point five zero five.

Taro: So, when you put that into a real system on a small board, you're dealing with a smaller physical footprint for the passive components while getting better performance metrics. That’s what matters for deployment outside of the lab, right?

Dev: It does matter for deployment. But I want to check the accuracy of their model against reality. They tested this under heavy load conditions, and they found that their theoretical model predictions agreed with the actual measurements within one point five percent when running a converter at a constant load resistance of twenty-five ohms.

Rosa: That level of agreement is pretty good for a model derived from complex topologies like this, especially when you're looking at high power density designs. It suggests the mathematical description holds up under stress.

Taro: So, to summarize what we've heard about this "Inductively Assisted N:one Switched-Capacitor Ladder Converter," it’s a topology that uses an inductor during demagnetization to cut average inductor current by twenty percent and boost output impedance by thirty percent compared to the standard SCLB <ref:2611.14778#pg1,Inductively Assisted N:1 Switched-Capacitor Ladder Converter>.

Dev: Exactly. It seems like it trades a slightly more complex switching sequence for better regulation and performance metrics across different conversion ratios, as shown by their closed-form model validation.

Rosa: So, to wrap up this discussion on the D-SCLB paper, it’s a design that aims to lower average current and improve impedance simultaneously using an inductively assisted switch in the bottom half of an N to one ladder buck converter.

Taro: It’s interesting how they manage the inductor current by letting those flying capacitors share the load during that demagnetization phase, which is key to achieving continuous regulation with M equal to D over N minus one plus D.

Dev: It certainly solves a problem in terms of managing the current flow, and the validation shows their model predictions are quite close to their experimental results under heavy load conditions.

Rosa: This paper suggests that this topology is directly extensible to higher conversion ratios and it's also suitable for monolithic integration into larger systems.

Taro: It’s a solid piece of work that looks like it could be used in real-world high-power applications where you need to manage current flow under tight constraints.

Dev: We'll wrap up our look at the D-SCLB here, and then we'll move on to another paper next.

The paper's summary: Rosa: So, to recap, this paper is about taking an N to one switched capacitor ladder buck converter and adding a tiny inductor during a specific part of the switching cycle to manage how much current flows through it.

Dev: Right, so they put this inductor in parallel with the capacitors while they are demagnetizing them, which basically means when you're running at that specific duty cycle M equal D over N minus one plus D, the average inductor current drops down to one minus M times the output current.

Rosa: Exactly. That lower average current is what lets them keep regulation going even when the system isn't running at its perfect native switched capacitor ratio. It makes it more continuous in practice.

Taro: Continuous regulation is key for robustness, but I’m wondering how this affects the physical design of these converters? Is it just a mathematical trick, or does it change how you build the actual hardware?

Dev: It changes the current flow management significantly, and they’ve done some real hardware testing. For a four by one converter, they claim this setup cuts down the average inductor current by up to twenty percent. That’s substantial for thermal management.

Rosa: And it also boosts the output impedance, which is a performance metric that shows how stable the circuit is when you push it hard under load conditions. They say it improves that by up to thirty percent compared to a standard switched capacitor ladder buck converter.

Taro: Thirty percent improvement in output impedance sounds like a big win for system stability. But the paper also talks about how they derived this output impedance model, saying it works for any N, whether you use optimal switch sizing or uniform switch sizing.

Dev: That modeling part is important because it gives designers a predictable performance number across different designs without having to run thousands of simulations just to check one specific case. It makes scaling easier.

Rosa: And they found that even when using uniform switch sizing, the improvement in output impedance is still around eight percent compared to the standard switched capacitor ladder buck converter. That's a consistent gain regardless of how you size your switches.

Taro: So they’ve got a mechanism that helps control current flow, improves regulation under specific duty cycles, and gives measurable performance gains on both current and impedance metrics across different design choices. But what about the real-world test results?

Dev: They validated the theoretical model against heavy load conditions, and they saw the predictions were within one point five percent of the actual measurements when running at a constant load resistance of twenty-five ohms. That shows their math is reliable in practice.

Rosa: It means this topology isn't just something you see on paper; it’s something that holds up when you run it under stress, which is exactly what we need for field applications. This suggests the D-SCLB is a viable path for high N to one conversions where current management is a major issue.

The paper's improvements: Rosa: So, we're looking at what this paper suggests is better than what we currently have by showing how to get more out of these N to one converters without making them overly complicated.

Dev: What they suggest is using that inductor not just for regulation, but also for preserving the optimal voltage and current stress on the main ladder capacitors. That’s a design consideration many people skip when they try to pack things down.

Rosa: Right, so it’s about managing the switch stress while simultaneously optimizing current delivery during that transition phase of demagnetization. It means you can push the system closer to its limits in terms of efficiency and power density.

Taro: If you’re pushing for high power density, reducing average inductor current by twenty percent is huge because less heat builds up in the surrounding components. That makes the whole package smaller too.

Dev: And it ties directly into the output impedance model they developed earlier, which shows a thirty percent improvement under optimal switch sizing. That means better stability when you're running those high-N ratios.

Rosa: It’s not just about making it work; it’s about making it perform better than the standard SCLB topology in specific ways, especially when you’re dealing with higher conversion ratios where current management gets tricky.

Taro: I wonder if this approach is scalable? Does this inductor trick still make sense if we move from a four by one to, say, an eight by one converter?

Dev: The paper says the topology is directly extensible to higher conversion ratios, which is a big deal for future designs because it’s not just a fix for one specific ratio. They also mention that it’s amenable to monolithic integration into larger systems.

Rosa: That means we could potentially shrink the overall passive component volume in a system because of this better current handling and the smaller inductor footprint they demonstrated in their prototypes.

Taro: So it moves beyond just fixing a single performance metric and gives us a whole new way to look at topology choice based on both regulation needs and physical constraints. That’s useful for autonomy systems too, where power efficiency is everything.

Conclusion: Rosa: So we’re wrapping up this look at the paper "An Inductively Assisted N:one Switched-Capacitor Ladder Converter." The main thing is that they found a way to use an inductor during a specific phase of operation to control average current and boost output impedance.

Dev: Exactly, and they showed that this method gives you measurable gains, like twenty percent less average inductor current for a four by one converter and thirty percent better output impedance under optimal switch sizing.

Rosa: It means we can design these N to one converters for higher power density while maintaining better stability across different conversion ratios.

Taro: I just think the way they handle the failure modes during that demagnetization phase is what really makes this interesting for autonomy, because it shows a graceful recovery path under heavy load.

Dev: That’s true, and their experimental validation showed that their theoretical model predictions were within one point five percent of the actual measurements when running at constant load resistance of twenty-five ohms.

Rosa: So they proved that this design holds up when you run it in the real world under tough conditions, which is what we need for field testing and deployment.

Taro: It does make sense that they focused on those specific duty cycles because that’s where the control loops get really tight.

Dev: And while they did show limitations on how much this inductor trick helps with non-optimal switch sizing, it still provides a solid framework for improving performance in many cases.

Rosa: Overall, this paper shows that by adding just one small component strategically, you can gain significant improvements in both efficiency and stability for these types of power converters.

Dev: It’s a good piece of work because it gives us a clear path forward for designing more robust N to one converters without having to completely redesign the core topology.

Taro: Next time we look at papers, we'll see how this kind of clever component placement might translate into something practical for more complex robotic power systems.

More episodes

← Home