Acronos P52 vs Cayin Jazz 80
Provisional comparison of the analog section. Reference system: Crossover A v2.4 (MR-A-01), 6 Ω nominal, Zmin 4.9 Ω at 1.4 kHz, 89 dB/W/m, listening at 3 m.
P52: 40 W nominal (C1/D, no THD condition) / 3× EPCOS B43645 610 µF / 450 V OBSERVED in silkscreen / Hongfa HF115F relay OBSERVED / blue Bourns trimmers OBSERVED on PCB-S (function ND) / Zout, P@1%, B+ ripple and bias = ND. Jazz 80: choke 1 H / 180 mA CONFIRMED / 26 W @3 %/8 Ω, 34 W @3 %/4 Ω and 46 W dynamic/3 Ω MEASURED-EXT in confirmed UL mode / consumption 148 W idle · 231 W max (measured) / individual bias per tube + VU-meters CONFIRMED. Thesis: if Zout ≤ 0.5 Ω + P@1 % ≥ 30 W + B+ ripple ≤ 100 mVpp + documented bias, the P52 is technically stronger for MR-A-01; without those measurements, the Jazz 80 retains the advantage of lower informational risk.
1. Decision question and thresholds
Decision question
For the reference system — Crossover A v2.4 MR-A-01, 6 Ω nominal / Zmin 4.9 Ω / 89 dB/W/m, listening at 3 m — under what technical conditions would the Acronos P52 be preferable to the Cayin Jazz 80?
| Threshold | Numeric criterion | Consequence if met |
|---|---|---|
| Zout / DF | Design target: Zout ≤ 0.5 Ω in the audible band (absolute loss over Zmin ≤ 1 dB). Minimum threshold: Zout ≤ 0.9 Ω (relative variation between 4.9 and 15 Ω ≤ 1 dB). ΔdB = 20·log10(Z/(Z+Zout)) | Stable response over the real impedance of the speaker |
| P@1% | ≥ 30 W @1% THD into 4–8 Ω | Positive margin in Scenario B (+3 dB room) |
| B+ ripple | ≤ 100 mVpp @ 100 Hz under load | SNR compatible with listening at 80 dB continuous |
| Bias and maintainability | Documented adjustment procedure / quartet replacement | Reasonable long-term operating cost |
2. Interior — visual comparison
| Element | P52 — Status | P52 — Source | Jazz 80 — Status | Jazz 80 — Source |
|---|---|---|---|---|
| HV capacitors | OBSERVED 3× EPCOS B43645 610 µF / 450 V (silkscreen B43645-S7617-M1 verified on all 3 units) | Direct reading + image 2026-10 | OBSERVED 3× Nichicon 330–470 µF / 85 °C (silver cross top) | Interior photo 2026-10 |
| Power relay | OBSERVED Hongfa HF115F (probable soft-start/inrush) | Image 2026-10 | OBSERVED Hongfa HF115F (probable soft-start/inrush) | Interior photo 2026-10 |
| Inductive filtering | PROBABLE by position · ND by function (two toroids, three hypotheses) | Physical position | CONFIRMED choke 1 H / 180 mA OBSERVED label in the image | Label + manufacturer |
| Coupling capacitors | OBSERVED 14× MKP CEC C2N 630-1000 V + 2 PIO | PCB-S photo | OBSERVED white Realcaps + red WIMAs | Interior photo 2026-10 |
| EMI input | DECLARED Schaffner FN9244B + 0.1 µF/1000 V OBS | Owner + photo | PARTIAL X caps + NTC + soft-start relay | Manufacturer |
| Rectifier | OBSERVED KBU1010 10 A | PCB-P photo | ND | — |
| Marked resistor | OBSERVED 10W10KJ (PCB-P) + 10K28 green (PCB-S) | PCB-P photo + image 2026-10 | — | — |
| Auxiliary regulation | PROBABLE MOSFET + TL431 + Bourns | PCB-P photo | ND | — |
| Transformers | OBSERVED 3× varnished EI (power + 2 outputs), metal partition | Video + image 2026-10 | OBSERVED 3× EI (center power supply, outputs at ends), full PCB | Interior photo 2026-10 |
| Supply rails | HYPOTHESIS three rails; PROBABLE single rail with local decoupling | Small electrolytics + blue trimmers on PCB-S; no visible series power resistors | ND | — |
3. Status of the evidence
The Cayin Jazz 80 has external laboratory measurements published by Stereo Magazin (via fairaudio / 7Review) that have been verified against primary sources:
- Power @3% THD / 8 Ω: 26 W MEASURED-EXT A2
- Power @3% THD / 4 Ω (4 Ω tap): 34 W MEASURED-EXT A2
- Complex music 8 Ω / 3 Ω: 33 W / 46 W MEASURED-EXT
- THD+N @ 1 W / 1 kHz: ≈ 0.18 % MEASURED-EXT · 0.2 % CONFIRMED (manufacturer)
- SNR: ≈ 100 dBA MEASURED-EXT · 90 dB A-weighted CONFIRMED (manufacturer)
- Idle / max consumption: 148 W / 231 W MEASURED-EXT
Measurement mode: the laboratory specifies that the measurements were performed in Ultralinear (UL) mode, which is the highest-power mode. The datum that appeared as ND in the previous comparison is now CONFIRMED (UL).
Official manufacturer specifications: Power Output TR: 20W+20W UL: 40W+40W · THD 0.2 % (1W/1kHz) · S/N 90 dB A-Weighted · Max Power Consumption 280 W · Net Weight 17 kg · Dimensions W×D×H 300×385×190 mm.
The Acronos P52 has no published measurements. The only available figures are the 40 W commercial nominal (without THD, frequency or load condition), classified as C1/D.
The 2026-10 images have made it possible to verify the EPCOS silkscreen on all three units of the HV bank and observe blue Bourns trimmers on PCB-S, which widens the range of hypotheses about the bias (four coexisting candidates) but does not close it.
| Datum | Acronos P52 | Cayin Jazz 80 | Note |
|---|---|---|---|
| P@3% THD / 8 Ω | ND | 26 W MEASURED-EXT A2 | Stereo Magazin, UL mode |
| P@3% THD / 4 Ω | ND | 34 W MEASURED-EXT A2 | 4 Ω tap, UL mode |
| P with complex music | ND | 33 W @8 Ω / 46 W @3 Ω MEASURED-EXT | Real dynamic capability |
| THD+N @ 1 W / 1 kHz | ND | ≈ 0.18 % MEASURED-EXT; 0.2 % CONFIRMED | THD+N, not pure THD |
| SNR | ND | ≈ 100 dBA MEASURED-EXT; 90 dB A CONFIRMED | Different sources |
| Idle / max consumption | ND | 148 / 231 W MEASURED-EXT · 280 W max CONFIRMED | New laboratory datum |
| B+ choke | ND (two toroids, function not determined) | 1 H / 180 mA CONFIRMED · OBSERVED label | Label + manufacturer + image |
| Bias / maintainability | ND (blue trimmers visible on PCB-S, function not determined) | Individual per tube + VU-meters CONFIRMED | Practical long-term advantage |
| Measurement mode (UL / triode) | ND | CONFIRMED UL (laboratory measurements) | Resolved in V3 |
| Zout / DF | ND | ND | Critical for both |
| NFB / %UL | ND | ND | Parameters that determine the sound |
| Nominal electrolytic temp. | 105 °C (B43645) OBSERVED | 85 °C (Nichicon) OBSERVED | See §5.1 |
| Transformers | 3× EI OBSERVED | 3× EI OBSERVED (center power supply, outputs at ends) | Constructive equality |
| Supply rails | HYPOTHESIS three rails; PROBABLE single rail | ND | See §5.1 |
4. Compared analog architecture
| Aspect | Acronos P52 | Cayin Jazz 80 KT88 | Assessment |
|---|---|---|---|
| Topology | PP ultralinear CONFIRMED (manufacturer); class of operation ND | PP ultralinear + switchable triode mode CONFIRMED | Jazz more flexible |
| Power tubes | 2× 6550 per channel (Tung-Sol OBS) | 2× KT88 per channel (EL34 compatible) | Secondary difference |
| Preamp / driver | 6Ж32П (EF86) + ECC803S | 2× 12AX7 + 2× 12AU7 | Different architectures |
| Phase inverter | ND until tracing | 12AU7 as driver/inverter CONFIRMED (manufacturer) | Jazz better documented |
| Coupling | 14× MKP CEC C2N 630-1000 V + 2 PIO OBS | White Realcaps + red WIMAs OBS | Both with quality capacitors |
| Bias | ND (blue trimmers visible on PCB-S, function not determined) | Individual per tube + VU-meters CONFIRMED | Jazz clearly superior in maintainability |
| OPT | EI + metal partition OBS | Cayin's own EI OBS | Both EI |
| Supply rails | HYPOTHESIS three rails; PROBABLE single rail | ND | Pending tracing on both |
The Acronos P52 uses an EF86 (6Ж32П) as the input tube, a low-noise high-gain pentode historically prized in phono preamps. This, combined with the ECC803S, describes a small-signal chain with a classic low-noise architecture. It is an architectural inference, not a measurement: the SNR of a power amplifier is dominated by the power supply, filaments and hum, not by the input tube.
The Cayin Jazz 80 uses 12AX7 in the preamp and 12AU7 as driver/inverter. Its external SNR measurements (≈ 100 dBA) indicate a quiet implementation; the detailed noise spectrum and the hum level at 50/100 Hz are not published.
The P52 and Jazz 80 use different small-signal architectures. In their 6550 and KT88 configurations, respectively, both employ a beam-tetrode push-pull output stage. The Jazz 80 also accepts EL34 as an alternative output tube. In the Jazz 80, Cayin documents the main function of its 12AX7 and 12AU7. In the P52, the presence of EF86 and ECC803S is identified, but the exact function of the ECC803S and the phase-inverter topology remain ND.
4.3.1 Inventory and equivalences
| Tube | Family | Functional equivalence | Status |
|---|---|---|---|
| 6Ж32П / EF86 (6267) | AF pentode | Published functional equivalence | CONFIRMED |
| ECC803S | Long-plate variant of the ECC83/12AX7 family | Nominal μ ≈100 | CONFIRMED for JJ Electronic |
| 12AX7 / ECC83 | Dual triode | Nominal μ ≈100 | CONFIRMED |
| 12AU7 / ECC82 | Dual triode | Nominal μ ≈20 | CONFIRMED |
| 6550 | Beam tetrode | — | CONFIRMED |
| KT88 | Beam tetrode | — | CONFIRMED |
| EL34 | Power pentode | — | CONFIRMED (alternative in Jazz 80) |
The ECC803S is described as a construction variant of the ECC83/12AX7 family, not as an absolute electrical equivalence with any 12AX7. Plate length is a documented construction characteristic; it does not by itself define electrical behavior in a specific circuit.
4.3.2 Function of each tube: documented, probable or ND
| Tube | P52 | Jazz 80 |
|---|---|---|
| 6Ж32П / EF86 | Input: PROBABLE | — |
| ECC803S | Gain / PI / driver: ND | — |
| 12AX7 | — | Preamp: PUBLISHED by Cayin |
| 12AU7 | — | PI + driver: PUBLISHED by Cayin |
| 6550 | PP UL power: CONFIRMED | — |
| KT88 | — | PP UL power: CONFIRMED |
| EL34 | — | PP UL power: CONFIRMED (alternative) |
PUBLISHED describes the function declared by the manufacturer; it does not imply that the exact internal topology of that function is documented. Cayin specifies that the 12AX7s act as preamp tubes and the 12AU7s as phase inverter and driver of the KT88s, but does not detail whether the inverter is a long-tailed pair, a cathodyne or another topology.
In the P52, the presence of a single EF86 per channel and its nature as a high-transconductance pentode point to it being used as an input or preamp stage. The ECC803S, with one tube per channel, can use its two triodes in various ways (gain + cathodyne, two triodes in LTP, gain + driver, etc.). These are general topological possibilities, not hypotheses assigned to the P52. The presence of blue trimmers on PCB-S does not confirm that they control the bias of the 6550s, so the exact biasing method remains ND.
4.3.3 EF86 / ECC803S: gain, family and microphonics
EF86: high potential voltage gain as a pentode; direct comparison with 12AX7: do not decide without a circuit. 12AX7: high-gain dual triode (μ≈100); in the Jazz 80 documented as preamp, with internal topology ND.
The EF86 was designed specifically for low-level audio and was used for decades as the input stage in high-fidelity preamps and microphone preamplifiers. Mullard introduced it in 1953 as "a low hum, low-microphony pentode", with a maximum hum level of 5 µV referred to the control grid under Mullard's specified test conditions, and with a reinforced electrode structure to reduce microphonics. That is a documented historical characteristic; it does not by itself imply a behavioral advantage in the specific P52.
Definition of microphonics. It is the conversion of mechanical vibrations of the tube into a spurious electrical signal that is incorporated into the circuit signal and can be subsequently amplified by the following stages. It is a problem of interaction between the mechanical structure of the tube, its mounting, the vibratory environment and the circuit gain, not a binary defect.
4.3.4 6550 vs KT88: electrical ratings
| Parameter | 6550 JJ | KT88 JJ |
|---|---|---|
| Maximum plate dissipation | 35 W | 42 W |
| Maximum plate voltage | 600 V | 800 V |
| Maximum screen voltage | 400 V | 600 V |
| Filament current | 1.6 A | 1.6 A |
| Socket | Octal | Octal |
The values correspond specifically to the JJ versions consulted; they should not be extrapolated as universal ratings. These ratings are characteristics of the tube. They do not by themselves determine the output power or the headroom of the complete amplifier. For that, one needs to know B+, screen voltage, quiescent current, plate-to-plate load of the OPT, %UL, maximum swing, NFB, bias, actual dissipation and transformer limit.
The Jazz 80 accepts KT88 or EL34 as declared by Cayin, with 40 W/channel UL for KT88 and 36 W/channel UL for EL34. This shows that the choice of power tube is part of an operating point designed by the manufacturer, not an absolute property of the amplifier.
5. Power supply, filtering and RF/EMI
| Aspect | Acronos P52 | Cayin Jazz 80 | Evidence status |
|---|---|---|---|
| Power transformer | EI OBSERVED | EI OBSERVED (center) | Same type in both |
| B+ choke | ND (two toroids, function not determined) | 1 H / 180 mA CONFIRMED · label OBSERVED | Jazz advantage |
| HV bank | 3× EPCOS B43645 610 µF / 450 V OBSERVED, 105 °C | 3× Nichicon 330–470 µF / 85 °C OBSERVED + choke | ≈ 1.83 mF only if in parallel ND |
| DC filtering | Two toroids → function ND (three hypotheses) | Choke confirmed CONFIRMED | P52: ND; Jazz: CONFIRMED |
| Auxiliary regulation | Probable (MOSFET + TL431 + Bourns) | Less documented | HYPOTHESIS P52 |
| Power relay | OBSERVED Hongfa HF115F | OBSERVED Hongfa HF115F | Same relay in both |
| Idle / max consumption | ND | 148 / 231 W MEASURED-EXT | New laboratory datum |
| Separate supply rails | HYPOTHESIS three rails; PROBABLE single rail | ND | See §5.1 |
| EMI input | Schaffner FN9244B / FN9280 module declared, not verified | Conventional filtering + X caps + NTC + soft-start relay | Both PARTIAL; neither verified |
| Physical separation | Metal partition power supply / PCB-S OBS | Compact layout on single PCB with all transformers in line | P52 with explicit separation |
The presence of a 1 H / 180 mA B+ filter choke is confirmed by the component label — clearly readable in the 2026-10 image as Cayin Electronic... 1H 180mA — and by the manufacturer's declaration. A choke in the supply reduces the 100 Hz ripple significantly compared to purely capacitive filtering, especially in the band where the dynamic impedance of the supply modulates the signal. Stereo Magazin explicitly confirms that the Jazz 80 delivers 46 W into 2 Ω, which it attributes to the delivery capability of the supply with choke.
If the bank of 3× 610 µF is actually in parallel (1.83 mF total), it would be a considerable energy bank for a 40 W stage. However, simple capacitance does not replace a choke in terms of 100 Hz ripple filtering. A large capacitive bank filters switching ripple well but presents a source output impedance that varies with demand; a choke introduces a high dynamic impedance that better isolates the output stage.
In an electrical network with pollution (harmonics, conducted RF, transients), a well-filtered EMI input is critical. The Schaffner module that the owner declares in the P52 is, in principle, a professional line filter; but the declaration is not verified. The Jazz 80, for its part, incorporates conventional filtering (X caps, NTC, relay) whose exact effectiveness is also not published.
6. Power, dynamic capability and load behavior
| Parameter | Acronos P52 | Cayin Jazz 80 (MEASURED-EXT) | Note |
|---|---|---|---|
| P@3% THD / 8 Ω | ND | 26 W A2 | Do not compare nominal vs @3% |
| P@3% THD / 4 Ω | ND | 34 W A2 | 4 Ω tap, UL mode |
| P complex music / 8 Ω | ND | 33 W | New datum, not in V2 |
| P complex music / 3 Ω | ND | 46 W | Real dynamic capability |
| THD+N @ 1 W | ND | ≈ 0.18 % | THD vs THD+N not equivalent |
| SNR | ND | ≈ 100 dBA | External source |
| Idle / max consumption | ND | 148 / 231 W | Measured in laboratory |
| P@1% (estimated) | 20 / 30 / 40 W (three cases) | ≈ 20 W / 8 Ω (range 18–24 W) EST | See §6.2 |
Interactive calculator
Adjust the three system parameters to obtain the required power per channel. Formula: P = 10^((SPL_target − SPL_1W@d − gain)/10), with SPL_1W@d = sens − 20·log10(d).
Reference model assuming a 5.3 Ω nominal speaker, useful as a sensitivity comparison with respect to the current model of the MR-A-01 (6 Ω nom / 4.9 Ω min).
| Scenario | Level @3 m / 95 dB Power / 40 W Margin | Acronos P52 (40 W nominal) | Cayin Jazz 80 (~20 W @1% est) |
|---|---|---|---|
| 89 dB @ 1 W/1 m, 3 m free field | 79.5 dB / 35.8 W / +0.5 dB | +0.5 dB if 40 W = P@1% (not demonstrated) | −2.5 dB @1% insufficient clean; @3% −1.4 dB at the limit |
| 89 dB @ 2.83 V/1 m over 5.3 Ω, 3 m | 79.5 dB / 54.1 W / −1.3 dB | −1.3 dB even if 40 W = P@1% → insufficient clean 95 dB free field | −4.3 dB @1% → insufficient; limit SPL or accept >1% THD |
| 89 dB SPL direct at 3 m with 1 W | 89 dB / 3.98 W / +10.0 dB | Wide margin +10.0 dB | Wide margin +7.0 dB @1% (20 W) |
Current reference system: Crossover A v2.4 MR-A-01. Sensitivity 89 dB/W/m CALCULATION HYPOTHESIS. Znom ~6 Ω; provisional Zmin 4.9 Ω @ ~1.4 kHz. EPDR estimated 3.0–3.5 Ω ND. Calculation: 20·log10(3) = 9.54 dB → 1 W @ 3 m = 79.46 dB per speaker. 95 dB mono → 35.8 W; ideal stereo +3 dB → 18.0 W/channel.
| Scenario A — 80/95 dB @3 m | Mono free | Stereo free /channel | Dry room /channel (+3+3 dB) | Live room /channel (+3+5 dB) |
|---|---|---|---|---|
| 80 dB continuous — 89 dB/W/m | 1.13 W | 0.57 W | 0.28 W | 0.18 W |
| 95 dB peak (80 + 15 dB crest) | 35.8 W | 18.0 W | 9.00 W | 5.68 W |
| Jazz ~20 W @1% est margin for 95 dB | −2.5 dB | +0.5 dB | +3.5 dB | +5.5 dB |
| P52 HYP 40 W @1% margin | +0.5 dB | +3.5 dB | +6.5 dB | +8.5 dB |
Calculation at 85/100 dB: 85 dB → 3.58 W mono; 100 dB → 113.3 W mono. Stereo applies +3 dB; the room adds +3 or +5 dB.
| Scenario B — 85/100 dB | Mono free | Stereo free /channel | Dry room /channel | Live room /channel |
|---|---|---|---|---|
| 85 dB continuous | 3.58 W | 1.80 W | 0.90 W | 0.57 W |
| 100 dB peak | 113.3 W | 56.8 W | 28.5 W | 18.0 W |
| Jazz 20 W @1% margin | −7.5 dB | −4.5 dB | −1.5 dB short at 1%; OK at 3% | +0.5 dB just |
| Jazz 26 W @3% / 34 W @4 Ω margin | −6.4 / −5.2 dB | −3.4 / −2.2 dB | −0.4 / +0.8 dB | +1.6 / +2.8 dB |
| P52 conservative case 20 W @1% | −7.5 dB | −4.5 dB | −1.5 dB | +0.5 dB |
| P52 medium case 30 W @1% | −5.8 dB | −2.8 dB | +0.2 dB | +2.2 dB |
| P52 optimistic case 40 W @1% | −4.5 dB | −1.5 dB | +1.5 dB | +3.5 dB |
Calculation at 90/105 dB: 90 dB → 11.33 W mono; 105 dB → 358.3 W mono. Monitoring levels, above typical domestic use.
| Scenario C — 90/105 dB | Mono free | Stereo free /channel | Dry room /channel | Live room /channel |
|---|---|---|---|---|
| 90 dB continuous | 11.3 W | 5.68 W | 2.85 W | 1.80 W |
| 105 dB peak | 358.3 W | 179.6 W | 90.0 W | 56.8 W |
| Jazz 20 W @1% margin | −12.5 dB | −9.5 dB | −6.5 dB | −4.5 dB |
| P52 optimistic 40 W @1% | −9.5 dB | −6.5 dB | −3.5 dB | −1.5 dB |
| Scenario | Continuous /channel (room +3/+5 dB) | Peak /channel (room +3/+5 dB) | Coverage at 1% | Typical use |
|---|---|---|---|---|
| A — 80/95 dB | 0.28 / 0.18 W | 9.00 / 5.68 W | Both more than sufficient | Normal domestic listening |
| B — 85/100 dB | 0.90 / 0.57 W | 28.5 / 18.0 W | Demanding — Jazz OK at 3% with 4 Ω tap; P52 covers in all three cases in live room | High-level listening |
| C — 90/105 dB | 2.85 / 1.80 W | 90.0 / 56.8 W | Neither at 1% | Monitoring — requires >3% THD or higher power |
7. Internal construction and isolation
| Aspect | Acronos P52 | Cayin Jazz 80 |
|---|---|---|
| Dedicated power / signal PCB | Observed (PCB-P + PCB-S separate) | Observed (specific PCBs for power supply + DAC and signal stage) |
| Power/signal separation | Explicit metal partition | Compact layout, no explicit partition |
| Point-to-point vs PCB | Mixed construction (PCB + twisted looms) | Entirely on PCB |
| Transformers | 3× varnished EI | 3× EI (center power supply, outputs at ends) |
| Bias access | Blue trimmers observed on PCB-S (function ND) | Individual per tube + VU-meters CONFIRMED |
| Maintainability | ND — no public documentation; no confirmed procedure | High — bias accessible per tube, VU-meters, public documentation |
| Output taps | ND — not observed in photos | 4 Ω and 8 Ω confirmed (gold-plated binding posts visible) |
| Consumption | ND | 148 W idle / 231 W max MEASURED-EXT · 280 W max CONFIRMED |
| Weight | ~40 kg PERSONAL COMMUNICATION, NOT VERIFIED | 16.8–17 kg (published) |
| Supply rails | HYPOTHESIS three rails; PROBABLE single rail with local decoupling | ND |
8. Verdict and measurements
| Criterion | Acronos P52 | Cayin Jazz 80 | Framing |
|---|---|---|---|
| Available objective evidence | Limited (visual inspection + catalogs) | Considerable (laboratory measurements in confirmed UL mode) | Jazz: lower informational risk |
| Architecture for this system | EF86 + possible C-L-C + possible regulation + metal partition — several NDs to resolve | Known, mature and documented architecture | Jazz: lower informational risk |
| Long-term maintainability | ND | High (individual bias per tube + VU-meters) | Jazz: lower informational risk |
| Load flexibility | ND (taps not observed) | 4 and 8 Ω taps confirmed | Jazz: lower informational risk |
| RF/EMI protection | Owner's declaration (not verified) | Conventional filtering + NTC + relay | Draw — insufficient evidence |
| Electrolytic thermal rating | 105 °C (B43645) OBSERVED | 85 °C (Nichicon) OBSERVED | P52: conditional ceiling |
| Purchase risk | High (many ND unknowns) | Low (documented and bounded unknowns) | Jazz: lower informational risk |
| Open questions that, if resolved favorably, would change the reading | Zout ≤ 0.5 Ω, P@1% ≥ 30 W, B+ ripple ≤ 100 mVpp, documented bias procedure, separate supply rails | Published Zout/DF | P52: higher conditional ceiling |
Reconciliation with MR-A-01 and three scenarios. In Scenario A (80/95 dB), both amplifiers are more than sufficient. In Scenario B (85/100 dB) in a dry room the Jazz 80 falls short at 1% but covers at 3% with the 4 Ω tap; the P52 covers from the medium case 30 W @1%. In a live room both cover. In Scenario C (90/105 dB), neither of the two covers the peak at 1% THD. The limiting magnitude in extreme scenarios becomes Zout/DF and current over Zmin 4.9 Ω.
The Cayin Jazz 80 is currently the better-characterized amplifier of the two. It has independent measurements of power (@3% THD in UL mode), distortion, dynamic capability with complex musical signal, idle/max consumption and visual confirmation of the 1 H / 180 mA choke. Its mature industrial PCB construction is documented. It is the known platform with fewer electrical unknowns.
The Acronos P52 presents, by inspection, an analog and power-supply architecture of interest: varnished EI transformers, physical power/signal separation via metal partition, dedicated PCB-P, industrial electrolytic bank with 105 °C rating. Under the hypothesis of 40 W/channel @ 1% THD / 4 Ω, it would have ≈ 3.5 dB of margin over the 95 dB peak at 3 m with two speakers. However, several details of its power supply remain as probable identification, owner's declaration or hypothesis until the circuit is traced and measured.
For this specific speaker, the critical pending magnitude is the P52's output impedance and its variation with frequency. A Zout ≤ 0.5 Ω, combined with 30–40 W @ 1%, low noise and a quiet supply under load, would turn the P52 into a technically strong proposition. Without those measurements, the Jazz 80 retains the advantage of lower informational risk.
| Framing | Amplifier |
|---|---|
| Lower informational risk | Cayin Jazz 80 |
| Higher conditional ceiling if hypotheses are confirmed | Acronos P52 |
| Fewer electrical unknowns to resolve | Cayin Jazz 80 |
| More open questions whose resolution would change the reading | Acronos P52 |
| Measurement | Method | Critical parameter |
|---|---|---|
| Zout | Voltage at no load and with known load, at 1 kHz and various frequencies | Zout = R_L · (V_no/V_with − 1) |
| DF | DF = Z_load / Zout | ΔdB = 20·log10(Z / (Z + Zout)). With Zout = 0.5 Ω and Z between 4.9 and 15 Ω → −0.8 to −0.3 dB |
| Speaker impedance and phase | Z(f) and phase sweep, with current probe and low-signal measurement | Real Zmin, real EPDR, Zmin frequency, minimum EPDR |
| THD+N vs power | At 4 Ω and 8 Ω, from 7 mW to 40 W, and with real load (RC network equivalent to Z(f) of the speaker) | P@1%, P@3%, knee point |
| FFT | At 7 mW, 1 W and 10 W | Noise and distortion profile |
| B+ ripple | Oscilloscope in AC coupled | Vpp at 100 Hz and harmonics |
| Frequency response | With real load (RC network reproducing Z(f) of the speaker) | ±dB variation in audible band |
| A-weighted noise | 20 Hz–20 kHz, input short-circuited | Residual level and 50/100 Hz spectrum |
| Stability | 1 kHz square wave + increasing capacitive load | Implied phase margin |
| Per-tap measurements | Repeat THD+N, Zout and response on each OPT tap | Identify optimal tap for 6 Ω |
| Clipping power | Increase power until visible clipping on oscilloscope | Real headroom over P@1% |
| Warming | Temperature of OPT, transformer, tubes and electrolytics after 30 min at 1/8 power | Real thermal derating |
| Channel balance | Gain and THD+N L vs R at 1 kHz | Symmetry of the two channels |
| %UL | Ratio of alternating signal between plate (pin 3) and g2 (pin 4) of the OPT | Real ultralinear percentage |
| NFB | Comparison of open-loop and closed-loop gain | Depth and bandwidth |
| Supply rails | Measure voltage and ripple at the B+ entry point to PCB-S and at the plate pins of EF86 and ECC803S, at rest and with signal | Number of rails, effective separation, residual modulation |
| Datum | Acronos P52 | Cayin Jazz 80 |
|---|---|---|
| Price | ND (artisanal build to order) | ND (consult dealer) |
| Warranty | ND | ND |
| Idle / max consumption | ND | 148 / 231 W MEASURED-EXT |
| Weight | ~40 kg PERSONAL COMM., NOT VERIFIED | 16.8–17 kg (published) |
| Dimensions | ND | 300×385×190 mm (W×D×H) |
| Tube cost | ND — 6550 quartet + EF86 + ECC803S | ND — KT88 quartet + 2× 12AX7 + 2× 12AU7 |
REFERENCES AND CREDITS
References and credits
Technical sources
- Stereo Magazin — independent measurements of the Cayin Jazz 80, in Ultralinear (UL) mode, via fairaudio / 7Review: 26 W @8 Ω, 34 W @4 Ω, 33 W @8 Ω complex music, 46 W @3 Ω complex music, THD+N 0.18% @1 W, SNR 100 dBA, consumption 148 W idle / 231 W max. (accessed 2026-10)
- Cayin — manufacturer documentation for the Jazz 80: functions of 12AX7 and 12AU7, 1 H / 180 mA choke, switchable UL/triode mode, individual bias per tube + VU-meters, max consumption 280 W, weight 16.8–17 kg, dimensions 300×385×190 mm, THD 0.2% (1 W/1 kHz), SNR 90 dB A-weighted. (accessed 2026-10)
- MusicTeck — description of the left VU-meter function as a control element for individual bias adjustment. (accessed 2026-10)
- TDK/EPCOS — Aluminum electrolytic capacitors, snap-in, B43645, 610 µF / 450 V. (accessed 2026-09)
- Hongfa — HF115F miniature high power relay. (accessed 2026-10)
- JJ Electronic — ECC803S, 6550, KT88. Datasheets. (accessed 2026-09)
- Mullard — EF86, Z729, CV2901, CV4085, historical documentation 1953. (historical document)
- Tube Amp Doctor (TAD) — EF86 / 6267 Premium Selected. Selection criteria. (accessed 2026-09)
- Schaffner — FN9244B / FN9280. (accessed 2026-09)
- Bourns — 3296 series. (accessed 2026-09)
- Texas Instruments — TL431. (accessed 2026-09)
Sources on the Acronos P52
- JC Valvular — manufacturing process video: ACRONOS P52 Amplifier — A work of art. How it was made.
- Manufacturer's commercial datasheet — 40 W PP UL with 6550 (no THD condition).
- Blog reader — personal communication: approximate weight of the P52 ~40 kg. Not verified.
- Interior images (2026-10) — provided by the owner of the unit: verify the B43645-S7617-M1 silkscreen on all three EPCOS units, identify the Hongfa HF115F relay, confirm the blue Bourns trimmers on PCB-S and document the arrangement of the 14 MKP CEC C2N, the green 10K28 resistor and the auxiliary electrolytics.
Sources on the Cayin Jazz 80
- Cayin — official product datasheet. (accessed 2026-10)
- Stereo Magazin / fairaudio / 7Review — laboratory measurements and analysis. (accessed 2026-10)
- MusicTeck — description of the bias procedure. (accessed 2026-10)
- Interior image (2026-10) — provided by the user: verifies the 1 H / 180 mA choke label, identifies the Hongfa HF115F relay, confirms the 3× large Nichicon capacitors, the Realcap and WIMA coupling capacitors, the 4 KT88/EL34 sockets and the 4 12AX7/12AU7 sockets, and shows the 3 EI transformers (power supply in the center, outputs at the ends).
Image credits
- Acronos P52 images — JC Valvular manufacturing video, reproduced for technical analysis purposes.
- Cayin Jazz 80 images — public manufacturer documentation and specialized press, reproduced for technical analysis purposes.
- Additional interior images (2026-10) of both amplifiers — provided by the owners of the units, reproduced for technical analysis purposes.
