Fezz Audio Titania MK2, 60 W push-pull valve amplifier
Technical report based on the official manuals for the integrated amplifier and its power stage, photographic analysis of the interior, reading of the board silkscreen, comparison with independent reviews, and explanation of the standardised method for measuring the damping factor. Manufactured in Poland by Fezz Audio with Toroidy toroidal transformers. EISA 2026-2027 award for Best Valve Amplifier.
Primary sources: manuals TITANIA-MK2-Integrated-ENG-230-V and Fezz-EVO-NEW-TITANIA-PA-MK2-ENG-110-230-V, Fezz Audio page, reviews from Stereolife, Audiodrom, HiFiNews and The Audio Tailor, OCR analysis of internal photographs, specialist distributor spec sheets and cross-verification of valve compatibility. Reference system: 35-year-old 3-way loudspeaker Crossover A v2.4 MR-A-01, 6 Ω nominal / 4.9 Ω minimum / 89 dB/W/m (calculation hypothesis, not measured), listening at 3 m.
What it is. A 60 W-per-channel push-pull valve integrated amplifier. It uses four KT88 valves in the power stage and three 12AX7s (ECC83) in the small-signal stage. It is manufactured in Poland, with Toroidy toroidal transformers. It won the EISA 2026-2027 award for Best Valve Amplifier. It accepts valves from the KT family: KT88 as standard and 6550 as a direct equivalent are confirmed; KT90, KT120 and KT150 are reported by distributors, pending confirmation from the manufacturer. It does not accept EL34, KT77, 6CA7, 6L6 or 5881.
The central unknown. Both official manuals declare a damping factor greater than 40. But there is no independent measurement of the MK2 that verifies this. The only available measurement in the family is from the power stage of the previous generation —Titania EVO Gen 1 (MK1), 45 W— and it yielded a value of 1. That figure is from another product and cannot be extrapolated to the MK2 without measuring it. See §5.2 and §5.3.
What is not known. The output impedance in ohms, the internal circuit topology, the core type of each transformer, and the exact electronic implementation of the Auto Bias.
GETTING STARTED — WITH NO PRIOR KNOWLEDGE (Basic guide G1–G2)
GETTING STARTED — WITH NO PRIOR KNOWLEDGE
A-level physics —logarithms, electrical power, impedance and series circuits— is enough to understand everything that follows. This guide covers two blocks: how a push-pull amplifier works and how reliable each data point is; and how power, impedance and damping relate. Technical details begin in Part I.
Basic guide
The idea that organises the whole report. The Fezz Audio Titania MK2 comes with two complete official manuals. In them the manufacturer declares very specific figures: 60 W per channel, distortion below 0.05 %, a damping factor greater than 40. Most of those figures are verifiable against the device. But there is one that is not, with the available documentation: the damping factor. The manuals declare it, but no independent party has measured it on the MK2. The only thing that exists is a measurement of the power stage of the previous generation, which gave a very different value. That is the open question of the report.
G1. How a push-pull works and how reliable each data point is
An amplifier converts the weak signal from the player —on the order of 1 V— into a signal with enough voltage and current to drive a speaker. Valve amplifiers have two blocks. The preamp increases voltage. The power stage supplies current. Everything is powered by a supply that produces a high DC voltage called B+.
Player → Preamp → Power stage → Output transformer → Speaker
(≈ 1 V) (voltage) (current) (impedance matching) (4–8 Ω)
↑
Power supply (B+)In the power stage there are two configurations. The first, the most common, is push-pull: two valves share the work and their currents combine in the output transformer. Advantages: DC components cancel in the transformer core (it does not saturate) and many even harmonics cancel each other (less distortion). The second, less frequent, is single-ended: a single valve per channel, with high linearity but low efficiency. The Acronos RS-1 in Appendix D is single-ended.
The output transformer matches impedances. If the turns ratio between primary and secondary is n, the impedance is multiplied by n²: with n = 27, an 8 Ω speaker is presented to the valves as 27² · 8 ≈ 5832 Ω.
The Titania MK2 uses four KT88s in push-pull class AB. Class AB means each valve conducts for slightly more than half a cycle. There are two variants. In AB1, the usual one in domestic audio, the grid never goes positive with respect to the cathode. In AB2, the grid can go positive at peaks, and that requires a driver stage capable of supplying current to the grid. AB1 is the most likely variant in the MK2 —the driver stage with 12AX7s can hardly supply the grid current that AB2 demands— but the manual does not specify it. PROBABLE
How reliable each data point is
In a technical report, statements of very different natures coexist. Some come from the manufacturer's official manual. Others we see in a photograph. Others result from applying a formula. And others are just reasonable intuitions. Mixing levels would be a mistake: the reader would end up giving the same credibility to a confirmed data point as to a suspicion. That is why every statement in the report carries an evidence label. The side aside (visible on desktop) shows the summary legend; Appendix A develops it.
| Label | What it means | Example in this report |
|---|---|---|
| CONFIRMED | Data published in the manufacturer's official manual. | 60 W per channel, THD < 0.05 %, 21.5 kg. |
| OBSERVED | Visible in the photographs or on the silkscreen. | Titania rev. 2/2025, the 220 µF/350 V capacitors. |
| MEASURED-EXT | Measurement published by an independent laboratory. May be of another product. | DF ≈ 1 measured by HiFiNews on the power stage Titania EVO Gen 1 (MK1), 45 W. Not the MK2. |
| PROBABLE | Reasonable physical inference, pending verification. | Location of the Auto Bias servo on the control sub-board. |
| CALCULATED | Comes from a formula based on known data. | Estimated Zout < 0.2 Ω if the declared DF > 40 were correct. |
| REPORTED-EXT | Data from a reliable third party, without own measurement or manufacturer confirmation. | Compatibility with KT90/KT120/KT150 according to distributors. |
| ND | Cannot be stated with current evidence. | Internal circuit topology, Zout in ohms of the MK2. |
G2. Power, impedance and damping
How many watts are needed. You must start with the speaker and the listening distance. The decibel (dB) is logarithmic: +10 dB multiplies power by 10, +3 dB by 2 (because 100.3 ≈ 1.995). The calculation follows four steps.
- Sensitivity. The reference speaker produces 89 dB with 1 W at 1 m. HYPOTHESIS of calculation from the Crossover A report, not measured.
- Distance. The level falls according to 20·log₁₀(d). At 3 m: 89 − 9.54 = 79.46 dB.
- Power. To reach a level L: P = 10^((L − 79.46)/10). For 100 dB: 113 W with a single speaker.
- Stereo. With two speakers and a decorrelated signal, +3 dB. Each channel contributes half.
With 60 W per channel, the Titania MK2 covers domestic listening with ample margin and 100 dB peaks with a tight margin. For 105 dB it falls short. The manual recommends using the 8 Ω tap with 6 Ω nominal speakers, not the 4 Ω tap.
Impedance and damping factor. A "6 Ω" speaker does not always measure 6 Ω: its impedance changes with frequency. The reference speaker drops to 4.9 Ω around 1.4 kHz and rises at the bass peak. An amplifier that behaves as a voltage source delivers more current when impedance drops. The output impedance (Zout) is a series resistance with the speaker. The damping factor summarises it: DF = Zload / Zout. The higher, the better the control.
The attenuation imposed by Zout follows the voltage divider formula, with 20·log₁₀ because we work with voltages:
With Zout = 0.2 Ω over 6 Ω: −0.29 dB. The variation between 4.9 Ω and 16 Ω is ≈ 0.24 dB, inaudible. The qualitative conclusion —excellent control, inaudible alteration— holds if the DF is indeed greater than 40.
| Damping factor | Estimated Zout (8 Ω) | Practical meaning |
|---|---|---|
| > 40 | < 0.2 Ω | Excellent control. The response follows the signal, not the speaker curve. |
| 10 – 40 | 0.2 – 0.8 Ω | Very good control. Spectral variation between 0.3 and 0.9 dB. |
| 3 – 10 | 0.8 – 2.7 Ω | Moderate control. Audible coloration possible. |
| < 3 | > 2.7 Ω | Low control. The response is clearly coloured. |
Quick exercises
- A speaker has 92 dB/W/m. What level does it produce with 1 W at 2 m?
- With the 89 dB/W/m speaker, what power is needed for a 100 dB peak at 3 m with a single speaker?
- If the Titania MK2 has DF > 40 into 8 Ω, what is its maximum Zout? And if the DF were 1, what would it be?
- With Zout = 0.2 Ω, what attenuation does the signal suffer over a 4.9 Ω speaker?
See solutions
- 92 − 20·log₁₀(2) = 92 − 6.02 ≈ 86 dB.
- 10^((100 − 79.46)/10) ≈ 113 W.
- DF > 40 → Zout < 0.2 Ω. DF = 1 → Zout ≈ 8 Ω.
- 20·log₁₀(4.9 / 5.1) ≈ −0.35 dB.
PART I — FEZZ AUDIO TITANIA MK2 REPORT
This report brings together Fezz Audio's two official manuals (the MK2 integrated and the PA MK2 power stage), the OCR of the interior, the EISA 2026-2027 award, the history of Toroidy with discrepancies between sources, power valve compatibility, the record of the declared damping factor and its pending independent verification, and the standardised method for measuring DF. Appendix D includes the reference case of the Acronos RS-1 SE KT150 as a methodological comparison.
0. Executive summary
20-second read
What it is. A 60 W-per-channel push-pull valve integrated amplifier, with four KT88s in the power stage and three 12AX7s (ECC83) in the small-signal stage. Manufactured in Poland by Fezz Audio, with Toroidy toroidal transformers. EISA 2026-2027 award for Best Valve Amplifier.
Valve family. KT88 (standard) and 6550 (direct equivalent) confirmed. KT90, KT120 and KT150 reported by distributors, pending manufacturer confirmation. Requires matched quartets. Details in Appendix B.
Pending verification. Both manuals declare DF > 40. No independent measurement of the MK2 exists. The only measurement in the family —power stage of the previous generation, Titania EVO Gen 1 (MK1), 45 W— yielded DF ≈ 1. The matter is developed in §5.2.
What is not known. Zout in ohms of the MK2, internal circuit topology, core type of each transformer, exact electronic implementation of the Auto Bias.
1. Method, evidence and limits
1.1 Approach and critical parameters
The reference listening system is the 35-year-old 3-way loudspeaker Crossover A v2.4 MR-A-01: 6 Ω nominal, 4.9 Ω minimum and 89 dB/W/m sensitivity. Listening is at 3 metres. The question guiding the report is simple: what does the Titania MK2 offer and how does it differ from other valve amplifiers?
The documentary advantage. The manufacturer publishes two complete manuals —the integrated and the power stage— and OCR analysis of the interior photographs has provided information that is not normally available: designer name, board revisions, visible components, fault LEDs, capacitor values. But there are no published independent measurements of the MK2's real performance at full power, and the only available measurement in the family —HiFiNews on the previous generation's power stage— corresponds to a different product.
1.2 Comparative master table
The following table summarises the most relevant parameters of the equipment with an indication of the level of evidence. Data from the official manual, OCR of the interior and board revisions have been incorporated.
| Parameter | Fezz Audio Titania MK2 |
|---|---|
| Configuration | Push-pull AB CONFIRMED · AB1 PROBABLE |
| Origin | Poland — Fezz Audio CONFIRMED |
| Award | EISA 2026-2027 Best Valve Amplifier CONFIRMED |
| Designer | Jakub Korpacz M.Sc.Eng OBSERVED |
| Main board revision | Titania rev. 2/2025 OBSERVED |
| Control sub-board | Universal Control System rev. 4/2023 OBSERVED |
| Standard power valves | 4× KT88 (2 per channel), JJ Electronic as standard CONFIRMED |
| Compatible KT family | KT88 and 6550 confirmed; KT90, KT120 and KT150 reported by distributors, pending manufacturer confirmation REPORTED-EXT — see Appendix B |
| Input/driver valves | 3× 12AX7 (ECC83) CONFIRMED |
| Transformers | Toroidal, by Toroidy CONFIRMED · core type ND |
| Nominal power | 60 W per channel CONFIRMED |
| Distortion (THD) | < 0.05 % at 1 kHz at full power CONFIRMED |
| Frequency response | 16 Hz – 120 kHz (−3 dB) CONFIRMED |
| Damping factor | Declared > 40 in both MK2 manuals CONFIRMED (manufacturer). No independent measurement of the MK2 ND. Measurement of the power stage EVO Gen 1 (MK1), 45 W: DF ≈ 1 MEASURED-EXT, another product — see §5.2 |
| Sensitivity | 0.7 V CONFIRMED |
| Input impedance | 50 kΩ CONFIRMED |
| Output impedance (Zout) | Not published. Estimated < 0.2 Ω if the declared DF > 40 were correct; ≈ 8 Ω if the MK2 behaved like the EVO Gen 1 (MK1) ND |
| Bias | Automatic (Auto Bias) CONFIRMED |
| Valve protection | Independent circuit for each KT88 with indicator LED CONFIRMED |
| Soft start | 2-min warm-up cycle with logo colour change CONFIRMED |
| Burn-in | 10 h preliminary; 40–60 h full; minimum 1 h per session CONFIRMED |
| Standby mode | Yes CONFIRMED |
| Inputs | 3× RCA + 1× Direct CONFIRMED |
| Outputs | 4 Ω / 8 Ω, Sub Out CONFIRMED |
| Recommended impedance | 6 Ω speakers → 8 Ω tap CONFIRMED (manual) |
| Optional modules (FEBS) | 2 slots: MM Phono, DAC, Bluetooth, XLR/RCA CONFIRMED |
| Remote control | Yes, infrared (volume, source, standby) CONFIRMED |
| Voltage selector | 230 V / 115 V selectable CONFIRMED |
| Main fuse | 3.15 A type T CONFIRMED |
| Signal ground switch | Yes, on/off on rear panel CONFIRMED |
| Coupling capacitors | Red film capacitors; Wima MKP10 PROBABLE |
| Filter capacitors | 8× 220 µF / 350 V OBSERVED · brand ND |
| Internal trimmers | 4–5 blue Bourns OBSERVED · exact function ND |
| Weight | 21.5 kg integrated · 21 kg PA MK2 power stage CONFIRMED |
| Dimensions | 420 × 380 × 200 mm CONFIRMED (official manual). Distributors publish 240 mm height REPORTED-EXT; it is probable PROBABLE that this corresponds to the height with the valve cage installed, but it is not confirmed by the manufacturer. |
| Consumption | 170 W at rest (manual); 160 W according to official website CONFLICTING DATA |
| Price | ≈ €3,750 base; FEBS modules ≈ €299 each REPORTED-EXT — indicative as of 2026-10, varies by distributor and country |
1.3 Evidence counter
It is worth distinguishing four situations that are sometimes confused under the generic label "not known": missing data (the manufacturer does not publish it, such as Zout in ohms), deduced data (inferred with uncertainty, such as compatibility with KT90/KT120/KT150), unknown data (there is no basis for inferring it, such as the exact internal topology) and data from another product (independent measurement of another device in the family, such as the DF of the EVO Gen 1 / MK1 power stage).
The master table in §1.2 has 37 rows. Some carry more than one label —"Transformers" carries CONFIRMED and ND; "Damping factor" carries CONFIRMED, ND and MEASURED-EXT—. The counter recounts the total number of labels present in those rows, not the number of rows.
| Label | Occurrences in §1.2 |
|---|---|
| CONFIRMED | 27 |
| OBSERVED | 5 |
| MEASURED-EXT | 1 (of another product: Titania EVO Gen 1 (MK1) power stage) |
| REPORTED-EXT | 3 (KT90/KT120/KT150 · dimensions with cage · price) |
| PROBABLE | 3 (Wima MKP10 · Dimensions · AB1) |
| ND | 5 (core, Zout, MK2 DF, capacitor brand, trimmer function) |
| Total labels | 44 |
| CONFLICTING DATA | 1 separate row (Consumption) |
The Titania MK2 is one of the best-documented devices in the series. The official manual provides most of the confirmed data. The OCR of the interior provides the observed data. The HiFiNews measurement is recorded as MEASURED-EXT but referred to another product (Titania EVO Gen 1 (MK1) power stage, 45 W), not the MK2. The real unknowns of the MK2 are reduced to Zout in ohms, the internal circuit topology, the core type of each transformer and the exact electronic implementation of the Auto Bias.
2. Identity and architecture
2.1 General view, silkscreen and designer
The Titania MK2 has a clean-lined chassis with a front panel that houses the power switch, input selector, volume control and IR receiver for the remote control. The toroidal transformers are housed in cylindrical cases inside the chassis. The device is available in several sheet-metal finishes.
Markings identified on the boards. The silkscreen of the main PCB reads «Titania rev. 2/2025», «Designed by Jakub Korpacz M.Sc.Eng», the reference TA6E-12 next to the upper terminal block and the label POWER AMPLIFIER ONLY near the right connector. Near the valve sockets the references KT88 and ECC83 appear repeatedly. In the bottom right corner an independent sub-board «Universal Control System rev. 4/2023» is also identified, also signed by the same designer, which houses the control logic.
General architecture
The Titania MK2 is a stereo push-pull class AB integrated amplifier. The power stage uses four KT88s, two per channel. The small-signal stage uses three 12AX7s (ECC83): two for the left and right channels, and a third that the manual explicitly designates as the "driver tube". The Toroidy toroidal transformers (one power and two output) are housed in cylindrical cases that also act as electromagnetic shielding.
The main signal PCB (rev. 2/2025) and the control sub-board (Universal Control System, rev. 4/2023) are connected by a ribbon cable. This physical separation between the signal path and the control logic is unusual in this segment and reinforces the modern vocation of the design.
2.2 Power and driver valves
The power stage uses four KT88s —beam tetrode, 42 W maximum plate dissipation, 800 V maximum voltage—, two per channel, with an individual fault indicator LED on each socket. The small-signal stage uses three 12AX7s (ECC83) —high-gain double triode (μ ≈ 100), low transconductance, widely used in audio—, also with a driver function for the output stage. Using the three 12AX7s in the same positions simplifies maintenance: only one valve type is needed for the entire small-signal section.
Fezz Audio uses JJ Electronic KT88s as standard. It offers an upgrade set consisting of Electro Harmonix KT88 and PSVane 12AX7-TII valves, sold together for about €200.
2.3 Titania MK2 block diagram
The following diagram summarises the signal chain, the power chain and the control subsystems of the Titania MK2 in a single stroke. The ND labels indicate unavailable parameters and the OBS labels indicate elements confirmed by OCR or by the manual. It is a functional architecture map, not a traced electrical schematic.
╔══ FEZZ AUDIO TITANIA MK2 · 60 W push-pull AB · Polish in-house design ══╗
║ Main board: Titania rev. 2/2025 · Designed by Jakub Korpacz M.Sc.Eng ║
║ EISA 2026-2027 award · Best Valve Amplifier ║
SIGNAL
[RCA ×3 / Direct In] → [12AX7 ×2] → [12AX7 ×1] → [4× KT88 PP] → [Toroidal output transformer] → [4/8 Ω speaker]
(obs: TO XLR prep) (preamp) (driver) (2 per channel) (Toroidy · parameters ND)
· 2 channels · 1 shared · Auto Bias (obs: TA6E-12 as
· 1 shared valve · 4 individual service label)
valve · ECC83 (obs) fault LEDs
· ECC83 (obs) · matched quartets
(obs) (manual)
CONFIRMED POWER VALVES: KT88 (standard) · 6550
REPORTED POWER VALVES: KT90 · KT120 · KT150
NON-COMPATIBLE POWER VALVES:
EL34 · KT77 · 6CA7 · 6L6 · 5881
(different family; the Silver Luna is the Fezz model for EL34)
MK1 → MK2 CONTRAST (HiFiNews):
MK1: 2× ECC83 drivers/inverters
MK2: 3× 12AX7 (2 preamp + 1 driver)
POWER SUPPLY
[AC] → [230/115 V switch] → [Toroidy toroidal transformer] → [SS rect. ND] → [8× 220µF/350V] → [B+ ND]
(manual · obs) (not valve) (obs) brand ND
├→ [3.15 A T fuse · manual]
├→ [Auto Bias · logic on UCS rev. 4/2023 sub-board]
├→ [Individual protection per KT88 · dedicated LED per valve · manual]
├→ [2-min warm-up cycle · logo red→white · manual]
└→ [Standby mode · manual]
CONTROL SUB-BOARD
[Universal Control System · rev. 4/2023 · Fezz Audio]
├→ [Auto Bias servo · calibration via 4–5 Bourns trimmers (obs)]
├→ [Individual electronic protection management per valve]
├→ [FEBS module switching control]
└→ [Start-up and warm-up cycle timing]
SERVICE AND ADJUSTMENTS
[Signal ground switch on/off · rear panel · manual]
[Test points TP2, TP3 · silkscreen · obs]
[Service labels POWER AMPLIFIER ONLY, TO XLR · obs]
EXPANSION (modular · 2 FEBS slots)
[Slot 1] → [MM Phono · DAC · Bluetooth · XLR/RCA] (≈ €299 per module)
[Slot 2] → [MM Phono · DAC · Bluetooth · XLR/RCA]
[Global NFB: ND] · [Class: AB (AB1 probable; manual/review)]
[Declared DF > 40 (both MK2 manuals) · NO INDEPENDENT MEASUREMENT OF THE MK2]
[DF ≈ 1 measured on the previous generation power stage (EVO Gen 1 / MK1, 45 W) · another product]
[THD < 0.05 % @ 1 kHz] · [Response: 16 Hz – 120 kHz (−3 dB)]
[60 W / channel] · [Mass: 21.5 kg]
[Burn-in 10 h preliminary / 40–60 h full · manual]Three observations from the diagram: Fezz Audio bets on vertical integration (three in-house toroidal transformers, with core type ND); the control architecture is physically separated from the signal path (Universal Control System rev. 4/2023 sub-board, with Auto Bias logic, per-valve protection and FEBS management); and the protection system is individual per valve, not global (each KT88 has its own fault LED).
The pure power stage version of the Titania —PA MK2, without preamp or FEBS, with 12 V trigger— is described in Appendix C.
3. Power supply
3.1 Toroidy toroidal transformers
The Titania MK2 power supply uses a toroidal transformer designed and manufactured by Toroidy, Fezz Audio's sister brand. The output transformers are also toroidal and the supplier is the same. Sources differ on the exact founding dates of Toroidy (HiFiNews: 1991; Stereolife: 1992) and on the debut of Fezz Audio as a brand (HiFiNews: 2000; Stereolife: Audio Video Show 2015). According to Stereolife Magazine, the design of the output transformers was carried out by Tomasz Lachowski, the founder's son. HiFiNews does not mention this detail.
3.2 Filter capacitors
The filtering capacity of the supply determines the level of ripple reaching the valves. The manual does not specify the brand, value or origin of the Titania MK2 filter capacitors.
- 8× 220 µF / 350 V in total OBSERVED
- Brand: ND (not legible in the available images)
- Possible distribution: 4 capacitors per channel, each channel in 2S2P configuration would give 220 µF effective @ 700 V per channel. HYPOTHESIS — the actual connection is not verified.
4. Control, bias and protection
4.1 Auto Bias
The Titania MK2 incorporates an automatic bias system (Auto Bias). Both manuals describe "automatic bias control" and add that it guarantees "maximum reliability and truly maintenance-free operation". The circuit continuously monitors the operating parameters of the valves and maintains the ideal bias without user intervention.
4.2 Protection and control architecture
The Titania MK2 brings together in its control subsystem three elements that are unusual in its category: individual protection for each power valve, FEBS modularity and control logic physically separated on the Universal Control System sub-board.
Individual protection per valve
The integrated amplifier's manual describes in section 7 ("Power tubes protection circuit"): "Every power tube has its own circuit protection system with separate light indicator". Each of the four KT88s has its own protection circuit and its own fault indicator LED. The documented procedure is:
- If the red LED lights up, switch off the amplifier.
- Check that the valves are correctly installed.
- Switch on again.
- If the red LED remains lit, the valve is defective and must be replaced.
The PA MK2 manual is more explicit about the technology of its version: it speaks of a "safety auto fuse" (resettable safety fuse) for each power valve. When a valve fails, the auto-fuse trips and the corresponding red LED lights up. The procedure is to switch off for 2 minutes, switch on again and, if it persists, replace the valve. Both systems fulfil the same objective with different philosophies. Which acts first or more effectively is not documented ND.
FEBS expansion modules
The Titania MK2 has two FEBS expansion slots (Fezz Extension Board System). The available modules are MM Phono, DAC, Bluetooth and XLR/RCA, at an indicative price of ≈ €299 each REPORTED-EXT. The PA MK2 power stage, by contrast, has no FEBS slots: its only expansion is XLR or RCA selection via a rear switch.
Control sub-board — Universal Control System
Photographic analysis of the interior has revealed the existence of an independent sub-board in the bottom right corner of the chassis, identified as "Fezz — Universal Control System — rev. 4/2023", also designed by Jakub Korpacz M.Sc.Eng. It is connected to the main board by a ribbon cable. It is the natural candidate to house the Auto Bias logic, the management of the electronic protection system, the switching of the FEBS modules and the warm-up cycle. The "Universal" reference suggests that it is used in several models of the Fezz range. The specific architecture of this sub-board (which ICs it carries, how it implements the bias servo, how it detects anomalies) remains ND.
5. Power, SPL and damping
5.1 Power and SPL
With 60 W per channel, the Titania MK2 offers an ample margin for domestic listening and a tight margin for high-intensity peaks. For the reference system (89 dB/W/m, listening at 3 m), the distance loss is about 9.5 dB.
| Scenario | Target peak SPL @3 m | W/ch required (+3 dB) | Margin over 60 W |
|---|---|---|---|
| Domestic | 95 dB | 18.0 W | +5.2 dB |
| Elevated | 100 dB | 56.8 W | +0.2 dB |
| Monitoring | 105 dB | 179.6 W | −4.8 dB |
SPL and required power calculator
Formulas: SPL_1W@d = sens − 20·log10(d) and P = 10^((target SPL − SPL_1W@d − factor)/10). Results per channel.
5.2 Damping factor (data without independent verification)
This section is the heart of the report. Everything else rests on it.
What the manufacturer declares. Fezz's two official manuals —the MK2 integrated and the PA MK2 power stage— declare a damping factor greater than 40. If that figure were correct, the output impedance would be below 0.2 Ω, which is an excellent value for a valve amplifier. The amplifier would control the speaker like a voltage source.
What has been measured. There is no independent measurement of the MK2 that allows that figure to be verified. What does exist is a measurement of the power stage of the previous generation. The HiFiNews review of the Titania EVO Gen 1 (MK1) power stage —45 W, January 2025— concluded that the 8 Ω and 4 Ω taps offer a source impedance approximately equal to the nominal impedance of each tap, that is, a DF ≈ 1. HiFiNews summarises it by pointing out that, under those conditions, the real response of the device is governed by the reactive impedance of the connected speaker, and recommends always using the 4 Ω tap, regardless of the speaker.
Why that measurement cannot be extrapolated. That measurement is from another product: the power stage, not the integrated, and from the previous generation, not the MK2. Fezz describes the MK2 topology as "completely new circuit topology" compared with the previous generation. The family clue is relevant —same manufacturer, same transformer supplier, same design philosophy— but it does not allow predicting the MK2's behaviour without measuring it.
The attenuation that Zout imposes on the signal is calculated with the voltage divider:
The following table shows the correctly calculated attenuation at 4.9 Ω and at 8 Ω for different hypothetical Zout values, including the two scenarios in play (DF > 40 and DF = 1):
| Hypothetical Zout | DF @8 Ω | Attenuation @4.9 Ω | Attenuation @8 Ω |
|---|---|---|---|
| 0.1 Ω | 80 | −0.18 dB | −0.11 dB |
| 0.2 Ω | 40 | −0.35 dB | −0.21 dB |
| 0.5 Ω | 16 | −0.84 dB | −0.53 dB |
| 1.0 Ω | 8 | −1.61 dB | −1.02 dB |
| 8.0 Ω | 1 | −8.41 dB | −6.02 dB |
What has been measured and what has not — the method
- With a purely resistive load of 8 Ω and 4 Ω, with sufficient power not to saturate.
- At several frequencies: at least 20 Hz, 100 Hz, 1 kHz, 10 kHz and 20 kHz.
- With an oscilloscope or True RMS multimeter to measure the voltage drop accurately.
- Noting the signal level of each measurement, because Zout can vary with excursion.
5.3 How DF is measured: standardised method
There is no single IEC or AES standard that defines "how to measure the damping factor". What exists is a de facto standard procedure for measuring output impedance (Zout), of which DF is a simple derivative. That general procedure is indeed covered in standards such as IEC 60268-3, which describes how to measure electrical parameters of amplifiers by the voltage-drop method. What the standard does not fix are the specific conditions —reference frequency, bandwidth, type of load— under which DF must be expressed as a catalogue figure.
5.3.1 The voltage-drop method
The method is known as the voltage-drop method or resistance insertion method (On-Off). It is the one used by laboratories such as HiFiNews and audio analysers such as Audio Precision. It consists of four steps:
- Measurement without load (Vnl). A signal of known frequency is injected —usually 1 kHz— and the output voltage is measured without connecting a speaker. It is advisable to use a minimum safety load (for example 100 Ω) instead of a pure open circuit.
- Measurement with load (Vl). A purely resistive load (typically 8 Ω or 4 Ω) is connected and the output voltage is measured again.
- Calculation of Zout. Zout = Rload · (Vnl − Vl) / Vl.
- Calculation of DF. DF = Rload / Zout.
Why it is preferable. The output transformer is inductive: measuring without load can generate overvoltages that damage the transformer or the valves. The two-load measurement is safer and is the variant used, for example, by JC Valvular (manufacturer of the Acronos RS-1).
5.3.2 Why conditions are not standardised
Although the method is standard, the exact conditions are not. That is where discrepancies between catalogue figures arise: the signal frequency (usually measured at 1 kHz to give the "official" figure, but Zout varies with frequency), the type of load (pure resistive load is standardised, which does not resemble the complex impedance of a real speaker) and the bandwidth (some manufacturers measure over a wide range, others only at 1 kHz).
The methodological parallel with other valve amplifiers in the same field is developed in Appendix D — Reference case: Acronos RS-1 SE KT150.
6. Operation, burn-in and safety
The manual describes the switch-on cycle precisely: first switch on the signal source, then the amplifier; turn the volume to minimum and switch on the Titania MK2. The device enters a 2-minute warm-up cycle. After 2 minutes, the Fezz logo on the front changes from red to white. Verify that the source is playing and gradually raise the volume. The switch-off sequence is the reverse. The PA MK2 follows the same procedure, with the exception that the volume is controlled by the external preamp.
Both manuals also specify the burn-in: about 10 hours of preliminary warm-up and between 40 and 60 hours of full burn-in, with minimum sessions of 1 hour to avoid short sessions during the initial period. The manuals justify it by "the presence of residual polarization of dielectric elements and due to a necessary stabilization of working parameters of the vacuum tubes".
The integrated amplifier's rear panel documents three service elements that are unusual in its category: 230 V / 115 V voltage selector, 3.15 A type T main fuse and signal ground on/off switch (to eliminate ground loops when several devices are integrated). The PA MK2 adds a 12 V trigger to automatically switch other system devices on and off. The manuals' safety warnings: do not switch on without valves or without speakers connected; do not open the cover (risk of high voltage even when disconnected); keep at least 10 cm of ventilation; replace fuses only with the original specification.
7. Assessment by axes: signal and control
This report assesses the Titania MK2 on two axes. Axis A (signal modernity) measures the degree to which the amplifier incorporates contemporary technical solutions for treating the musical signal. Axis B (control modernity) measures how it manages its own operation: monitoring, protections, maintenance and connectivity.
| Axis | Element | Fezz Titania MK2 | Assessment |
|---|---|---|---|
| A | Input/driver stage | 3× 12AX7 (ECC83) CONFIRMED | Classic, proven solution |
| A | Coupling capacitors | Red film capacitors; Wima MKP10 PROBABLE | High-quality polypropylene |
| A | Output transformers | Toroidy toroidal CONFIRMED | Low emission, high stability |
| A | Topology | Push-pull AB (AB1 probable), redesigned (rev. 2/2025) CONFIRMED | Redesigned signal path |
| A | Power valves | 4× KT88 standard · KT family compatible CONFIRMED | Robust and flexible beam tetrode |
| A | Rectification | Solid state (not valve) PROBABLE | Modern solution |
| B | Bias | Automatic (Auto Bias) with internal calibration points CONFIRMED | Low maintenance for the user |
| B | Valve protection | 4 independent circuits with LED per KT88 CONFIRMED | Superior granularity |
| B | Valve compatibility | KT88, 6550 confirmed; KT90, KT120, KT150 reported REPORTED-EXT | Sonic flexibility with caveats |
| B | Control sub-board | Universal Control System rev. 4/2023 OBSERVED | Physical logic/signal separation |
| B | Soft start | 2-min warm-up cycle with logo colour change CONFIRMED | Modern |
| B | Standby mode | Yes CONFIRMED | Modern |
| B | Remote control | Infrared: volume, source, standby CONFIRMED | Modern |
| B | Expansion modules | 2 FEBS slots (Phono, DAC, BT, XLR/RCA) CONFIRMED | Modularity uncommon in valve integrateds |
| B | Ground switch | Signal ground on/off on rear panel CONFIRMED | Unusual function |
| B | Voltage selector | 230 V / 115 V CONFIRMED | Standard in the international market |
| B | 12 V trigger | Only on the PA MK2 power stage (not on the integrated) CONFIRMED | Integration in automated systems |
| B | Documentation | Complete official manual + identifying silkscreen OBSERVED | Very good |
| B | Distribution network | Established distributors | Good support |
8. Closing
8.1 Balance
What we know.
- The device is officially documented: both manufacturer manuals publish all essential specifications.
- EISA 2026-2027 award for Best Valve Amplifier.
- The OCR of the interior provides traceability: designer (Jakub Korpacz M.Sc.Eng), board revisions and filter capacitor values (8× 220 µF / 350 V).
- Excellent control architecture: Auto Bias, individual protection per KT88, FEBS modularity, remote control, standby and ground switch.
- Compatible with the standard KT family (KT88, 6550 confirmed; KT90, KT120 and KT150 reported by distributors). Requires matched quartets. Not compatible with EL34, KT77, 6CA7, 6L6 or 5881 (for EL34 there is the Silver Luna).
- Uses Toroidy toroidal transformers.
- THD < 0.05 % at 1 kHz at full power: an excellent figure for a valve amplifier.
- Its FEBS modularity is uncommon in valve integrated amplifiers of its category: it allows adding a DAC, phono, Bluetooth or XLR/RCA without changing equipment.
- A sister power stage exists (PA MK2) with the same valve complement, same power and 12 V trigger, without preamp or FEBS.
What we do not know or is pending.
- MK2 damping factor. Both manuals declare more than 40, but there is no independent measurement of the MK2. The only available measurement in the family is from the previous generation power stage (Titania EVO Gen 1 (MK1), DF ≈ 1). It is the most critical data point pending verification. See §5.2.
- MK2 Zout in ohms. It depends on resolving the previous point: if the declared DF were correct, it would be below 0.2 Ω; if the MK2 behaved like the EVO Gen 1 (MK1) power stage, it would be approximately 8 Ω.
- Internal circuit topology. Phase splitter, driver, KT88 driver, NFB topology.
- Core type of each transformer.
- Exact electronic implementation of Auto Bias.
- Real behaviour of KT90/KT120/KT150 in the MK2. Distributors report them as compatible, but the manufacturer does not confirm it in writing.
8.2 Quick decision table
| If your priority is… | Guidance |
|---|---|
| Automation and low maintenance | Titania MK2 — Auto Bias and individual protection per valve |
| Easy diagnosis of valve faults | Titania MK2 — 4 individual LEDs per KT88 |
| Sonic flexibility with the standard KT family (KT88, 6550) | Titania MK2 — confirmed by manual |
| Experimenting with KT90/KT120/KT150 | Titania MK2 with caution — reported by distributors, requires manufacturer confirmation |
| Sound with EL34 / KT77 | Fezz Audio Silver Luna — not compatible with Titania MK2 |
| Modularity and future connectivity | Titania MK2 integrated — two FEBS slots |
| Integration in automated systems | Titania PA MK2 — 12 V trigger |
| Integration in complex systems (integrated) | Titania MK2 — signal ground switch |
| International use | Both — 230/115 V selector |
| Speaker control with low Zout | Pending verification of the MK2's real DF — if the declared DF is correct, good control; if the MK2 behaved like the EVO Gen 1 (MK1) power stage (DF ≈ 1), the choice of speaker becomes critical |
| Complete official documentation | Both — exhaustive manuals |
| Highest quality transformers | Both — Toroidy toroidal |
8.3 What CANNOT be concluded from this report
This list serves to prevent reading more than what is there. With only manuals, photos and a measurement of the previous generation, none of the following can be stated:
- That the MK2's real DF is greater than 40. There is no independent measurement of the MK2.
- That the MK2's Zout is below 0.2 Ω. It is a conditional estimate.
- That the MK2 behaves like the EVO Gen 1 (MK1) power stage (DF ≈ 1). The measurement is from another product, from another generation.
- That Auto Bias eliminates all maintenance. The manuals require matched quartets and replacement procedures.
- That changing from KT88 to KT120 or KT150 objectively improves the sound. Without controlled listening.
- That KT90/KT120/KT150 are fully compatible. Reported by distributors, not confirmed by the manufacturer in writing.
- That the transformer core type is toroidal. The external shape is cylindrical, but visual confirmation of the core is ND.
- That the distortion curve of the MK2 at different powers is that of the THD < 0.05 % at 1 kHz at full power. That manual data point is specific; it does not describe behaviour at intermediate powers.
- That the reference speaker's sensitivity is 89 dB/W/m. It is a calculation hypothesis from the Crossover A report, not a measurement.
- That the MK2's global NFB is high, low or non-existent. It is not documented.
- That the electronic implementation of Auto Bias is of a specific kind. The Universal Control System rev. 4/2023 sub-board is visible, but its internal architecture is ND.
APPENDICES
Supporting material. Appendix A explains how to interpret the evidence labels. Appendices B and C expand on compatible valves and the sister power stage. Appendix D develops the parallel case of the Acronos RS-1. Appendix E details the pending measurement protocol. Appendix F gathers the sources and credits.
APPENDIX A — HOW TO READ THE EVIDENCE LABELS
Appendix A — How to read the evidence labels
Each statement in the report carries an evidence label indicating its level of support. This appendix develops the summary legend from the side aside (visible on desktop) for the mobile and print versions, where the aside is hidden.
The eight labels, one by one
| Label | What it means | What would be needed to upgrade it |
|---|---|---|
| CONFIRMED | Data established by official manufacturer documentation (manual) or verified silkscreen. | It is already at the highest level. It can only be degraded if a documentary contradiction appears. |
| OBSERVED | Directly visible in photographs or on the board silkscreen. | Confirm the interpretation with the manufacturer, or measure the component directly. |
| CALCULATED | Mathematical result obtained by applying a formula to known data. | Experimentally verify the result with an instrument. |
| PROBABLE | Reasonable physical inference, coherent with what is observed, but unverified. | Direct measurement or documentary confirmation. |
| HYPOTHESIS | Possibility coherent with the available data, but not sufficiently supported. | Additional evidence to turn it into PROBABLE or discard it. |
| REPORTED-EXT | Data provided by a reliable third party (distributor, specialist publication), without own measurement or manufacturer confirmation. | Official manufacturer confirmation or independent measurement. |
| MEASURED-EXT | Measurement published by an independent laboratory, with identifiable conditions. May refer to another product. | Measurement of the product under report under the same conditions. |
| ND | Not determinable with current evidence. | Direct measurement, manufacturer schematic or additional documentation. |
How several labels combine in a single row
A single statement may carry several labels at once, because different parts of the sentence have different levels of evidence. The damping factor row in the master table of §1.2 reads:
Each part of the sentence responds to a different source. A reader who reads all three parts with the same label will not understand the real state of the question. A reader who distinguishes the three levels will see clearly that the question remains open and will know what type of measurement would close the matter.
APPENDIX B — POWER VALVE COMPATIBILITY
Appendix B — Power valve compatibility
The Titania MK2 does not work only with KT88s. The manufacturer optimises the design for that model, but the amplifier accepts other valves from the same electrical family. This appendix gathers what is documented, what specialist distributors report and the practical conditions that the user must respect.
B.1 Compatible valves
| Valve | Type | Status in the MK2 | Notes |
|---|---|---|---|
| KT88 | Beam tetrode | STANDARD | Reference valve. JJ Electronic as standard; upgrade with Electro Harmonix KT88 + PSVane 12AX7-TII. |
| 6550 | Beam tetrode | COMPATIBLE | American cousin of the KT88. Very similar electrical characteristics. Requires matched quartets. |
| KT90 | Beam tetrode | REPORTED-EXT | Evolution of the KT88 with higher dissipation. Auto Bias re-adjusts automatically. Heater consumption similar to KT88. Not confirmed by the manufacturer. |
| KT120 | Beam tetrode | REPORTED-EXT | More power. Heater consumption 1.7–1.95 A. Verify with the manufacturer. |
| KT150 | Beam tetrode | REPORTED-EXT | Maximum power option. Heater consumption ∼2 A. Requires explicit confirmation from the manufacturer. |
B.2 NON-compatible valves
| Valve | Type | Status in the MK2 | Notes |
|---|---|---|---|
| EL34 | Pentode | NOT COMPATIBLE PROBABLE | Different electrical family. Lower power and plate voltage. The Fezz model for EL34 is the Silver Luna (2×35 W, 4× EL34). |
| KT77 | Beam tetrode | NOT COMPATIBLE PROBABLE | Direct substitute for the EL34. Same limitations. |
| 6CA7 | Pentode | NOT COMPATIBLE PROBABLE | American equivalent of the EL34. Same limitations. |
| 6L6 / 5881 | Beam tetrode | NOT COMPATIBLE PROBABLE | Different electrical family (lower voltage and power than the KT88s). The MK2 manual does not mention them explicitly, but by electrical family they are not compatible. |
- Plate voltage: EL34/KT77 and 6L6/5881 handle much lower voltages than the KT88s. The MK2 circuit is designed for KT88 voltages.
- Dissipation: EL34 25 W; KT77 32 W; 6L6 30 W; KT88 42 W. The MK2 operating point is calibrated for KT88.
- Auto Bias: its adjustment range is designed for the KT88 family. It cannot compensate for the curve differences of an EL34 or a 6L6.
- Consequence: attempting to use any of these valves in the MK2 can cause unstable operation, damage to the valves or the output transformer, and is not covered by the warranty.
B.3 The role of Auto Bias
The Titania MK2's Auto Bias system is what makes changing valves in the KT family viable without manual intervention: when replacing the quartet, the circuit automatically readjusts the bias point. This does not exempt compliance with the matching requirement that the manual demands, but it does eliminate the need to manually recalibrate the bias after each change.
- Detects the new quiescent current of each valve.
- Adjusts the negative grid voltage to return it to the optimum operating point.
- Corrects drift during use (valve ageing).
B.4 Practical conditions for valve changes
| Condition | Why it matters |
|---|---|
| Matched quartets | Explicit manual requirement. Without prior matching, Auto Bias may compensate poorly and some valve could operate outside its safe point. |
| Verify heater consumption | KT88 and 6550 consume 1.6 A; KT120 can reach 1.95 A; KT150 approximately 2 A. Four of these valves plus the three 12AX7s (≈ 0.3 A each) imply a total consumption of around 7.3 A with KT88 versus 8.9 A with KT150 (+22 %). It is advisable to confirm with the manufacturer that the transformer has margin. |
| Confirm with Fezz Audio or the distributor | For KT120 and KT150, it is advisable to ensure in writing that the power transformer and the rest of the circuit operate within their safety margins. |
| Preamp valves: 12AX7 / ECC83 | For the small-signal section, the manual requires replacement with the same type or direct equivalent. |
| Burn-in after the change | The manual recommends 40–60 h burn-in for complete sound stabilisation. |
- With previously matched quartets.
- After confirming with Fezz Audio or the authorised distributor the specific electrical compatibility.
- Verifying that the total heater consumption does not exceed the transformer's capacity.
- Respecting the manual's safety procedures: disconnect from the mains before touching valves, wait for them to cool, do not open the cover.
APPENDIX C — THE SISTER POWER STAGE: TITANIA PA MK2
Appendix C — The sister power stage: Titania PA MK2
Fezz Audio also markets a version of the Titania as a pure power stage, without an integrated preamp. It is called the Titania Power Amplifier (PA) MK2 and shares most of the electrical architecture with the integrated, but not all of it. It is the same MK2 generation of 60 W as the integrated; it must not be confused with the Titania EVO Gen 1 (MK1) 45 W power stage, which is the previous generation.
C.1 Differences with respect to the integrated
| Feature | Titania MK2 integrated | Titania PA MK2 stage |
|---|---|---|
| Inputs | 3× RCA + 1× Direct | 1× RCA + 1× XLR (switchable) |
| Volume control | Yes, integrated in the chassis | No; controlled by the external preamp |
| FEBS modules | 2 slots | No |
| 12 V trigger | No | Yes, input/output |
| Weight | 21.5 kg | 21 kg |
| Valves | 4× KT88 + 3× 12AX7 | 4× KT88 + 3× 12AX7 |
| Power | 60 W | 60 W |
| THD | < 0.05 % at 1 kHz full power | < 0.05 % at 1 kHz full power |
| Bandwidth | 16 Hz – 120 kHz | 16 Hz – 120 kHz |
| Declared DF | > 40 | > 40 |
| Zin | 50 kΩ | 50 kΩ |
| Sensitivity | 0.7 V | 0.7 V |
| Consumption | 170 W / 270 W | 170 W / 270 W |
| 230/115 V switch | Yes | Yes |
| Signal ground switch | Yes | Yes (indicated as "GND audio loop switch") |
| Warm-up | 2 min, logo red→white | 2 min anode voltage delay |
C.2 Why it matters for this report
The PA MK2 power stage is part of the same MK2 generation of 60 W as the integrated. The fact that both manuals declare the same DF greater than 40 indicates that it is a coherent specification within the MK2 family, not a one-off figure from a single document. However, there is still no independent measurement of the MK2 —integrated or stage— that verifies that value. The only independent measurement available in the family is that of the Titania EVO Gen 1 (MK1) 45 W power stage, carried out by HiFiNews in January 2025, which gave DF ≈ 1. It is not possible to speak of a direct contradiction between the manufacturer and the measurement, because they refer to different generations of the product, and the measurement is only of the power stage. What exists is an unverified manufacturer declaration and a family precedent that advises caution.
C.3 Protection system: interesting difference
Both products protect each valve independently, but with different technologies: the integrated documents an electronic protection circuit, while the stage documents individual resettable auto-fuses. Both manuals record the same number of LEDs (four), one for each KT88. Which of the two technologies acts first or more effectively is not documented ND.
APPENDIX D — REFERENCE CASE: ACRONOS RS-1 SE KT150
Appendix D — Reference case: Acronos RS-1 SE KT150
The Fezz Titania case is not unique in the valve audio landscape. A Spanish product in the same field, the Acronos RS-1, is a good example of how the declared DF figure can be subject to the same caution. The full RS-1 report, with NFB modelling, the class A ceiling and the manufacturer's testimony, is published on Nauscopio.
D.1 What the Acronos RS-1 is
The Acronos RS-1 is a single-ended (SE) KT150 valve amplifier, with a single valve per channel. It belongs to a different topological family from the Fezz (which is push-pull). Its architecture is that of the pure SET, with the advantages and disadvantages that this entails: high intrinsic linearity, absence of crossover distortion, but low efficiency and, typically, high Zout.
D.2 Empirical precedents in SE KT150
Commercial amplifiers with KT150 in single-ended tend to have high Zout. Two documented precedents, one with independent measurement and another with a figure declared by the manufacturer:
| Model | Power | Global NFB | Reported DF |
|---|---|---|---|
| Icon Audio ST30SE | 28 W UL | Low-feedback design | ≈ 6 (Hi-Fi World) · DF ≈ 7–17 depending on mode, in HiFiNews measurements |
| Dared Saturn Signature | 25 W | −39 dB global | ≈ 6 (declared) — internally inconsistent: with −39 dB of NFB and an open-loop Zout of 5–6 Ω, the DF should be around 130, not 6. One of the two data points does not mean what it appears to. |
D.3 The figure published by Et Sona and its traceability
The figure that circulates for the Acronos RS-1 —DF ≈ 50 and THD 0.01 % @ 18 W— comes from the Et Sona review, not from an official JC Valvular spec sheet. This has an important consequence for the analysis: it is not recorded that the manufacturer publishes them in its official spec sheet.
D.4 What can and cannot be concluded
| Statement | Status |
|---|---|
| The RS-1 is an SE KT150 with one valve per channel | OBSERVED |
| Documented SETs with KT150 have low DF, between 6 and 17 depending on mode and publication | MEASURED-EXT (one) + REPORTED-EXT (the other) |
| Et Sona publishes DF ≈ 50 for the RS-1 | REPORTED-EXT |
| Et Sona's measurement methodology is known | ND |
| The RS-1 has deep-loop global NFB | ND |
D.5 Parallel between the two cases
Both the Fezz Titania MK2 and the Acronos RS-1 are European products in the same field, with pure valve architectures. In both cases, the declared DF figure is not supported by an independent measurement of the product itself:
| Aspect | Fezz Titania MK2 | Acronos RS-1 |
|---|---|---|
| Topology | Push-pull KT88 | Single-ended KT150 |
| Declared DF | > 40 (official MK2 manual) | ≈ 50 (Et Sona, not manufacturer) |
| Source of the figure | Manufacturer (two MK2 manuals) | External review |
| Independent measurement of the current product | None recorded for the MK2. Yes for the EVO Gen 1 (MK1) power stage (DF ≈ 1, HiFiNews). | No published independent measurement |
APPENDIX E — PENDING MEASUREMENT PROTOCOL
Appendix E — Pending measurement protocol
To close the unknowns of this report —and, above all, to verify the MK2's declared damping factor— a series of measurements on a physical unit would be needed. These are the five priority ones:
| Priority | Measurement | Purpose |
|---|---|---|
| 1 | Zout by the resistive divider method, on each tap (4 Ω and 8 Ω), at several frequencies (20 Hz – 20 kHz) | Verify the MK2's declared DF (> 40): check whether the MK2 behaves as a voltage source (high DF) or whether it follows the pattern of the previous generation power stage (DF ≈ 1). |
| 2 | Square-wave step response of the OPT over an 8 Ω load | Reveals primary inductance, leakage inductance and high-band behaviour. |
| 3 | Distortion curve versus power | Establishes P@1 % THD, a data point that no manual publishes. |
| 4 | Verification of the connection of the 8 capacitors (distribution per channel and exact topology) | Confirms the real effective capacity of the filter bank per channel. |
| 5 | Consultation with the manufacturer about DF, transformer margins and full KT120/KT150 compatibility | Confirm in writing the DF figure, the admissible heater consumption and the compatibility of the valves reported by distributors. |
Desirable complementary measurements: macro photo of the Universal Control System rev. 4/2023 sub-board (to identify the ICs of the Auto Bias and protection system); supply ripple with oscilloscope; and direct comparison between Titania MK2 integrated and PA MK2 stage on the same speaker.
APPENDIX F — REFERENCES AND CREDITS
Appendix F — References and credits
Primary sources
- Fezz Audio · official manual for the MK2 integrated amplifier. TITANIA-MK2-Integrated-ENG-230-V. Official PDF document in English, 8 pages. Authoritative source for all documented specifications and functions of the integrated. (accessed 2026-10)
- Fezz Audio · official manual for the PA MK2 power stage. Fezz-EVO-NEW-TITANIA-PA-MK2-ENG-110-230-V. Official PDF document in English, 8 pages. Authoritative source for the PA MK2 specifications and the resettable auto-fuse protection system. (accessed 2026-10)
- Fezz Audio · official website. fezzaudio.com. (accessed 2026-10)
- EISA 2026-2027 award. EISA (Expert Imaging and Sound Association). eisa-awards.org. Official quote: «The Titania MK2 significantly advances Fezz's KT88-based integrated tube amplifier, delivering up to 60Wpc of power to partner confidently with a wider range of loudspeakers.» (accessed 2026-10)
- HiFiNews · review of the Titania EVO Gen 1 (MK1) power stage, January 2025. hifinews.com — Fezz Audio Sagita Prestige/Titania pre/power amp. The unit measured by HiFiNews is the power stage of the Titania EVO Gen 1 (MK1) —45 W, previous generation of the family—, not the Titania MK2 that is the subject of this report. The review documents an output impedance of 7.4–8.0 Ω on the 8 Ω tap, over the 20 Hz – 20 kHz range, corresponding to a DF ≈ 1. It also documents that the MK1 used two ECC83s as drivers/inverters and that it already had auto-bias. The MK2 report cites the review as a family precedent, with the methodological caution explained in §5.3.2. (accessed 2026-10)
- History of Toroidy and Fezz Audio, with discrepancies between sources. HiFiNews places the founding of Toroidy in 1991 and the entry of Fezz Audio into the market in 2000. Stereolife Magazine, article "From transformers to tube amplifiers: the story of Fezz Audio" (April 2026), places it "probably" in 1992 and dates the debut of Fezz as a brand at the Audio Video Show in 2015. The data on the re-adaptation of Biazet machinery, the role of Tomasz Lachowski in the design of the output transformers and the anecdotal origin of the name "Fezz" come only from Stereolife Magazine; HiFiNews does not mention them. (accessed 2026-10)
- Fezz Audio Titania MK2 · photographic analysis. High-resolution photographs of the interior. OCR source: Titania rev. 2/2025, Designed by Jakub Korpacz M.Sc.Eng, Universal Control System rev. 4/2023, TA6E-12, KT88, ECC83, POWER AMPLIFIER ONLY, TO XLR, TP2, TP3. Capacitor reading source: 8× 220 µF / 350 V. (accessed 2026-10)
- Specialist distributor spec sheets. Commercial documentation confirming compatibility with the KT family (6550, KT90, KT120, KT150) and the Silver Luna model for EL34. Also the source of the 240 mm height figure, complementary to the 200 mm in the official manual. (accessed 2026-10)
Complementary sources
- Titania MK2 reviews. Stereolife, Audiodrom, The Audio Tailor, Hifi-voice. The Audiodrom review rates the sound at 98 % and highlights "ultra-high-endová kultivovanost".
- Valve datasheets. JJ Electronic (KT88, 12AX7), Genalex/New Sensor (KT88 Gold Lion), Electro Harmonix (KT88, KT90, KT120, KT150), PSVane (12AX7).
- Historic reference design. Mullard Ltd. Circuits for Audio Amplifiers (Booklet TP413).
- Reference system. Crossover A v2.4 MR-A-01, 6 Ω nominal / 4.9 Ω minimum / 89 dB/W/m (calculation hypothesis, not measured).
- Cross-reference comparative report. Acronos P52 vs Cayin Jazz 90 (nauscopio, V3.3, October 2026).
- DF measurement methodology. IEC 60268-3 "Sound system equipment - Part 3: Amplifiers" describes the general procedure for measuring electrical parameters, including output impedance by the voltage-drop method. However, the standard does not fix specific conditions (frequency, bandwidth, type of load) under which DF must be expressed as a catalogue figure, which explains the disparity of conditions between manufacturers and laboratories. HiFiNews, Audio Precision (instrument manufacturer) and others apply the method with different conditions.
- Acronos RS-1 reference case. The full RS-1 report is published at nauscopio.blogspot.com/p/acronos-rs-1-single-ended-kt150-en.html. It includes loop-gain modelling, the class A ceiling calculation, the manufacturer's testimony on DF calculation, the list of the 12 factors that determine DF and a 16-measurement verification protocol. The DF ≈ 50 and THD 0.01 % @ 18 W figure comes from the Et Sona review, not from an official manufacturer spec sheet. The empirical precedents of the Icon Audio ST30SE and Dared Saturn Signature are at low DF values. For the Icon Audio ST30SE, Hi-Fi World published DF ≈ 6 and HiFiNews' own laboratory gave different figures (DF between 7 and 17 depending on mode), which illustrates that even two measurements of the same device can differ according to conditions.
- JC Valvular testimony on DF calculation. The designer of JC Valvular —manufacturer of the Acronos RS-1— has explained, in the context of the RS-1 report, that in his case DF is set as an objective in the output transformer design based on 12 factors, and that the commercial figure is a verified average at 1 kHz with fixed resistive loads. (2026-10)
- Two-load method for measuring Zout. Variant of the voltage-drop method used by some manufacturers to avoid measuring in open circuit, given the inductive nature of the output transformer. Formula: Zout = R₁·R₂·(V₁−V₂) / (R₁·V₂ − R₂·V₁).
