Half Pound Filter (HPF) for Cryogenic Attenuation

• Half Pound Filter (HPF) is a cryogenic copper powder filter that uses the skin effect to attenuate high-frequency microwave and radio signals.
• The conventional design embeds a signal conductor in copper powder, achieving strong multi-GHz attenuation but introducing significant parasitic capacitance.
• Recent multilayer designs interpose a low-permittivity gap to drastically reduce capacitance while preserving effective microwave absorption.

Searching arXiv for the specified topic and cited papers to ground the article in recent literature. A half pound filter (HPF) is a cryogenic copper powder filter used in low-temperature wiring to provide strong, broadband attenuation of microwave and radio-frequency radiation that would otherwise propagate down measurement lines and heat a sample. In its classic form, an HPF consists of a straight signal conductor routed through a grounded metal housing densely packed with metal powder and a binder, typically copper powder in a copper chassis. Its operating principle is the skin effect: as frequency increases, the skin depth decreases, current is confined to a progressively smaller effective cross-section in the powder medium, and dissipation rises accordingly. Conventional HPFs are valued as “last-mile” absorbers in quantum transport and related cryogenic experiments, but their monolithic filled geometry also creates substantial parasitic capacitance to ground, which can severely limit low-frequency and MHz-band measurement bandwidth. Recent work has therefore recast the HPF geometry as a multilayer powder filter that preserves GHz attenuation while greatly reducing capacitance to the grounded chassis (Pradhan et al., 31 Jul 2025).

1. Terminology and scope

In cryogenic wiring, a “Half Pound Filter” is the workhorse copper powder filter: a straight signal conductor passes through a metal housing filled with metal powder and binder, and the roughly half-pound of copper powder provides a very large metal surface area surrounding the conductor. A classic HPF scales this geometry to a longer conductor, a larger block, and roughly half a pound of copper powder intimately packed around the wire inside a copper housing, yielding tens to well over 100 dB attenuation across the multi-GHz band when built with centimeter-scale lengths and large fill volumes (Pradhan et al., 31 Jul 2025).

The acronym “HPF” is not unique across the literature. In astronomical time-series analysis, HPF denotes “High-pass Filter” in the “High-pass Filter Periodogram,” explicitly not “Half Pound Filter” (Albentosa-Ruiz et al., 2024). In Herschel PACS data reduction, HPF likewise denotes a running-median “High-Pass Filter,” again not “Half Pound Filter” (Popesso et al., 2012). A separate 2026 animation paper also uses the name “Half Pound Filter” for a modification of the 1 Euro Filter in real-time animation blending, and that paper does not provide an explanation for the name (Lasagno, 25 Feb 2026). In cryogenic instrumentation, however, “HPF” conventionally refers to the powder-filter architecture.

This terminological overlap suggests that disambiguation is essential when HPF appears outside the immediate context of cryogenic transport lines.

2. Physical principle and transmission-line description

The attenuation mechanism of the cryogenic HPF is rooted in the skin effect. At microwave frequencies, currents in the metal powder are confined to a skin depth

$$\delta(\omega) = \sqrt{\frac{2}{\omega \mu \sigma}},$$

which shrinks with frequency. Because the effective cross-section for current flow in the powder becomes small and lossy, high-frequency radiation is strongly dissipated before reaching the sample (Pradhan et al., 31 Jul 2025).

Powder filters are distributed, lossy transmission lines. A simple model assigns per-unit-length parameters $R(\omega)$, $L$, $C$, and $G$, with propagation constant

$$\gamma(\omega) = \alpha(\omega) + j\beta(\omega) = \sqrt{(R + j\omega L)(G + j\omega C)}.$$

In these filters, $G \approx 0$. At high frequency, the effective series resistance per unit length rises approximately as $\sqrt{\omega}$ because $\delta \propto 1/\sqrt{\omega}$, so one may write $$R(\omega) \propto 1/\delta(\omega) \propto \sqrt{\omega \mu/\sigma}$$. When $R(\omega)$0 dominates and $R(\omega)$1 is modest, the attenuation constant scales approximately as

$R(\omega)$2

The experimentally accessible attenuation is reported as

$R(\omega)$3

This distributed-line viewpoint is important because the HPF is not merely a lumped absorber. Its effectiveness depends on the proximity of lossy powder to the current-carrying conductor, the total length of the dissipative path, and the electromagnetic coupling to the grounded housing (Pradhan et al., 31 Jul 2025).

3. Conventional architecture and its principal limitation

The canonical HPF geometry embeds the signal line directly in a conductive, high-$R(\omega)$4 powder/binder medium inside a grounded copper chassis. In the small test article used by Pradhan et al., the conventional comparator consisted of a 0.25 mm diameter copper wire running straight through an oxygen-free copper cavity of internal dimensions $R(\omega)$5, with $R(\omega)$6, 20, or 40 mm, fully filled with the commercial metal-powder epoxy Eccosorb CR-124 and terminated with SMA launchers. This was described as the canonical HPF geometry scaled to a small test article (Pradhan et al., 31 Jul 2025).

The Achilles’ heel of this geometry is parasitic capacitance to ground. Because the signal line is intimately embedded in a grounded, conductive, high-permittivity medium, the structure forms a substantial stray capacitance $R(\omega)$7 to the chassis. Large HPFs routinely measure many nF of capacitance to ground, with $R(\omega)$8 on the order of nF/m in conventional filled geometries. That capacitance forms an RC shunt with the source or device resistance and sets a corner frequency

$R(\omega)$9

For typical device or source resistances in the 10–100 k$L$0 range, the RC pole can be pushed into the 10–100 kHz regime, which limits usable measurement bandwidth, increases phase lag in transimpedance amplifiers, and can even promote sample heating by providing a low-impedance return for out-of-band noise (Pradhan et al., 31 Jul 2025).

A concise comparison is given below.

Feature	Conventional HPF-like geometry	Multilayer geometry
Signal environment	Wire embedded directly in CR-124 filling the cavity	Wire inside CR-124-filled polymer tube
Chassis coupling	Continuous capacitive path to grounded chassis	Low-$L$1 gap between tube and chassis
Capacitance slope	About 2 nF/m	Approximately 50 pF/m
Total capacitance	nF-level for large filled blocks	Less than 10 pF for complete parts
High-frequency attenuation	$L$2 dB above $L$3 GHz for 20 mm test build	30–40 dB above $L$4 GHz for 20 mm; $L$5 dB for 40 mm

The central trade-off of the traditional HPF is therefore clear: maximization of GHz absorption by direct embedding of the conductor in the lossy medium simultaneously maximizes the unwanted capacitive path to ground.

4. Multilayer reformulation of the HPF

The multilayer powder filter reported by Pradhan et al. retains the beneficial GHz absorption of a powder filter while breaking the direct capacitive path to the grounded chassis. The key geometric change is to separate the powder surrounding the signal wire from the grounded copper housing with a low-$L$6 gap (Pradhan et al., 31 Jul 2025).

In the reported implementation, the same 0.25 mm copper wire is threaded through a thin polymer tube with inner diameter 2.2 mm and wall thickness 0.25 mm; the tube is completely filled with CR-124. Separately, the inner walls and lid of the copper chassis are coated with approximately 0.5 mm of CR-124. The tube assembly is then installed in the chassis so that a typically 1.5 mm radial air gap remains between the tube’s outer surface and the CR-124-coated chassis. The wire ends are soldered to the SMA pins as usual (Pradhan et al., 31 Jul 2025).

This multilayer configuration places the lossy powder exactly where it is most effective for skin-effect attenuation—proximate to the current-carrying conductor—while interposing a low-permittivity gap between the powder/wire region and the grounded housing. For simple parallel-plate or coax-like gaps, one may estimate capacitance per unit length as $L$7 or, more crudely, $L$8; increasing $L$9 or reducing $C$0 lowers $C$1. The measured multilayer capacitance slope matches an air-gap model with $C$2 for a $C$3 mm gap, which is consistent with the intended decoupling mechanism (Pradhan et al., 31 Jul 2025).

A plausible implication is that the multilayer design should be understood not as a rejection of the HPF principle, but as an architectural refinement of it: the lossy medium is confined to the region that controls $C$4, while the region that controls shunt capacitance is deliberately made low-permittivity.

5. Measured performance and consequences for cryogenic readout

Quantitatively, the multilayer design reduces parasitic capacitance by a factor of approximately 40. In measurements of capacitance versus length, the conventional HPF-like filters showed a linear slope of about 2 nF/m, whereas the multilayer design showed approximately 50 pF/m. For the multilayer parts, the absolute capacitance was so small that the SMA connectors dominated, and the total capacitance was less than 10 pF (Pradhan et al., 31 Jul 2025).

The characteristic impedance of the multilayer filter was approximately 65 $C$5, slightly higher than 50 $C$6 because of reduced capacitance per unit length. The paper states that this mismatch is unproblematic in cryogenic transport lines operating in the kHz–MHz regime where matching is not critical (Pradhan et al., 31 Jul 2025).

RF transmission measurements showed the expected attenuation trade-off. At room temperature and $C$7 mm, the conventional HPF-like build provided very strong attenuation, $C$8 dB above approximately 7 GHz, whereas the multilayer build provided broadband 30–40 dB attenuation above approximately 10 GHz. For the multilayer geometry, attenuation increased with length, reaching approximately 60 dB for $C$9 mm across the multi-GHz band. The attenuation was essentially unchanged at 4.2 K, indicating robust cryogenic performance. Below approximately 100 MHz, attenuation from the skin effect was negligible, so baseband signals passed with minimal insertion loss (Pradhan et al., 31 Jul 2025).

The most consequential system-level result concerns measurement bandwidth and noise. In a representative cryogenic transimpedance amplifier measurement of a 100 k$G$0 device with $G$1 k$G$2, flat band approximately 2 MHz, and input-referred current noise approximately $G$3, the multilayer filter, adding approximately 5 pF, preserved the flat transimpedance magnitude and phase to well above 1 MHz. By contrast, the conventional filter, adding approximately 50 pF, lowered the bandwidth by about an order of magnitude and elevated the noise floor above approximately 300 kHz. The paper identifies this contrast as a direct manifestation of the RC limit set by $G$4 at the amplifier input (Pradhan et al., 31 Jul 2025).

These measurements establish the essential trade: relative to a conventional HPF of similar length, the multilayer design sacrifices some top-end attenuation in exchange for dramatically higher usable baseband bandwidth and cleaner phase response.

6. Fabrication practice, system integration, and comparison with other filters

The reported builds used an OFHC copper chassis, SMA launchers, a straight 0.25 mm diameter copper wire, and Eccosorb CR-124 as the powder medium. The copper block provides thermalization and a grounded shield. Practical guidance given in the paper includes maintaining a continuous lossy sleeve around the conductor by completely filling the inner tube, preserving an approximately 1.5 mm low-$G$5 gap to the chassis, coating rather than filling the chassis with approximately 0.5 mm CR-124, and starting with block lengths of 20–40 mm when MHz-band operation matters. Based on the measured spectra, a 20 mm multilayer block yields approximately 30–40 dB attenuation above 10 GHz, while a 40 mm block reaches approximately 60 dB. Two 20–40 mm blocks at successive cold stages can be used if more attenuation is needed while keeping capacitance low and distributed (Pradhan et al., 31 Jul 2025).

The same source notes several implementation cautions. Conductive epoxy bridges across the air gap can substantially raise capacitance. The epoxy should be degassed before filling, voids should be avoided, and curing should follow the manufacturer’s schedule for vacuum and cryogenic compatibility. Thermalization is improved by bolting the copper housing to the dilution refrigerator cold plate with a good thermal interface. Because CR-124 is magnetically loaded, it should be mounted away from field-sensitive regions if necessary (Pradhan et al., 31 Jul 2025).

Relative to other cryogenic filters, the multilayer HPF occupies a specific niche. Compared with a conventional single-body HPF, it reduces capacitance by approximately 40 times, slightly increases characteristic impedance, and trades some attenuation for low capacitance. Compared with Thermocoax or mineral-insulated coax, it is presented as a complementary final absorber at the coldest stage. Compared with absorptive Eccosorb blocks in 50 $G$6 housings, it can approach their attenuation while minimizing added capacitance for high-impedance sample lines. Compared with discrete RC filters, powder filters remain superior for microwave photon suppression at GHz frequencies (Pradhan et al., 31 Jul 2025).

7. Limitations, misconceptions, and present significance

A common misconception is that maximizing attenuation alone defines HPF quality. The recent cryogenic literature shows that this is incomplete: a filter can excel as a GHz absorber and still degrade the measurement chain by introducing large shunt capacitance. In that sense, the defining weakness of the traditional half pound filter is intrinsic to its filled geometry rather than incidental to poor construction (Pradhan et al., 31 Jul 2025).

The multilayer reformulation also has limits. Maximum RF power handling is set by absorber heating and thermal sinking into the copper chassis; the paper does not specify a power limit and notes that the small mass of CR-124 in compact blocks implies modest power limits compared with a classic half-pound HPF. Attenuation can be further increased by lengthening the filter or cascading stages, but doing so raises DC resistance slightly and can raise capacitance if the low-$G$7 gap is not preserved. Powder type and grain size affect the high-frequency loss through $G$8 and $G$9, and higher-$$\gamma(\omega) = \alpha(\omega) + j\beta(\omega) = \sqrt{(R + j\omega L)(G + j\omega C)}.$$0 gap materials degrade the capacitance advantage (Pradhan et al., 31 Jul 2025).

In present usage, the HPF remains a foundational concept in cryogenic measurement wiring: bury the line in metal powder to exploit the skin effect and suppress microwave leakage. The contemporary significance of the design lies in the recognition that this concept need not be tied to a monolithic filled block. The multilayer design shows that the HPF’s attenuation mechanism can be retained while its most consequential systems-level penalty—large parasitic capacitance—can be sharply reduced. For quantum transport and sensing setups that require both GHz photon suppression and MHz-band readout integrity, this reconfiguration represents a direct continuation of the half pound filter tradition rather than a departure from it (Pradhan et al., 31 Jul 2025).